Theory of Abel Grassmann's Groupoids

Educational Publisher
Columbus ▌ 2015
Madad Khan ▌Florentin Smarandache ▌Saima Anis
Theory of Abel Grassmann's Groupoids
Peer Reviewers:
Prof. Rajesh Singh, School of Statistics, DAVV, Indore (M.P.), India.
Dr. Linfan Mao, Academy of Mathematics and Systems,

Chinese Academy of Sciences, Beijing 100190, P. R. China.
Mumtaz Ali, Department of Mathematics, Quaid-i-Azam University,

Islamabad, 44000, Pakistan
Said Broumi, University of Hassan II Mohammedia,

Hay El Baraka Ben M'sik, Casablanca B. P. 7951, Morocco.
Educational Publisher
Columbus ▌2015
Copyright: © Publisher, Madad Khan1, Florentin Smarandache2, Saima Anis1. 2015
1
Department of Mathematics, COMSATS Institute of Information Technology, Abbottabad, Pakistan
2
Department of Mathematics & Science, University of New Mexico, Gallup, New Mexico, USA
[email protected]
The Educational Publisher, Inc.

1313 Chesapeake Ave.
Columbus, Ohio 43212, USA
Toll Free: 1-866-880-5373
www.edupublisher.com/
ISBN 978-1-59973-347-0
Contents
Preface 7
1 Congruences on Inverse AG-groupoids 11

1.1 AG-groupoids 11
1.2 Inverse AG -groupoids 13
2 Structural Properties of -AG -groupoids 21

2.1 Gamma Ideals in -AG-groupoids 21
2.2 Locally Associative -AG -groupoids 37
2.3 Decomposition to Archimedean Locally Associative AG-subgroupoids 44
3 Embedding and Direct Product of AG-groupoids 47

3.1 Embedding in AG-groupoids 47
3.2 Main Results 48
3.3 Direct Products in AG-groupoids 50
4 Ideals in Abel-Grassmann’s Groupoids 61

4.1 Preliminaries 61
4.2 Quasi-ideals of Intra-regular Abel-Grassmann's Groupoids 64
4.3 Characterizations of Ideals in Intra-regular AG-groupoids 75
4.4 Characterizations of Intra-regular AG-groupoids 84
4.5 Characterizations of Intra-regular AG -groupoids 93
5 Some Characterizations of Strongly Regular AG-groupoids 101

5.1 Regularities in AG-groupoids 101
5.2 Some Characterizations of Strongly Regular AG-groupoids 102
6 Fuzzy Ideals in Abel-Grassmann's Groupoids 109

6.1 Inverses in AG-groupoids 111
6.2 Fuzzy Semiprime Ideals 114
7 (2; 2 _q) and (2; 2 _qk)-fuzzy Bi-ideals of AG-groupoids 119

7.1 Characterizations of Intra-regular AG-groupoids 123
7.2 (2; 2 _qk)-fuzzy Ideals of Abel-Grassmann's 132
7.3 Main results 132
7.4 Regular AG-groupoids 138
8 Interval Valued Fuzzy Ideals of AG-groupoids 151

8.1 Basics 151
8.2 Main Results using Interval-valued Generalized Fuzzy Ideals 153
9 Generalized Fuzzy Ideals of Abel-Grassmann's Groupoids 163
9.1 (2 ; 2 _q )-fuzzy Ideals of AG-groupoids 163
9.2 (2 ; 2 _q )-fuzzy Quasi-ideals of AG-groupoids 177
10 On Fuzzy Soft Intra-regular Abel-Grassmann's Groupoids 191
10.1 Some Characterizations Using Generalized Fuzzy Soft Bi-ideals 199

10.2 References 202
Theory of Abel Grassman's Groupoids 7
Preface
It is common knowledge that common models with their limited boundaries
of truth and falsehood are not su¢ cient to detect the reality so there is a
need to discover other systems which are able to address the daily life
problems. In every branch of science problems arise which abound with
uncertainties and impaction. Some of these problems are related to human
life, some others are subjective while others are objective and classical
methods are not su¢ cient to solve such problems because they can not
handle various ambiguities involved. To overcome this problem, Zadeh [67]
introduced the concept of a fuzzy set which provides a useful mathematical
tool for describing the behavior of systems that are either too complex or are
ill-de…ned to admit precise mathematical analysis by classical methods. The
literature in fuzzy set and neutrosophic set theories is rapidly expanding
and application of this concept can be seen in a variety of disciplines such as
arti…cial intelligence, computer science, control engineering, expert systems,
operating research, management science, and robotics.
Zadeh introduced the degree of membership of an element with respect
to a set in 1965, Atanassov introduced the degree of non-membership
in 1986, and Smarandache introduced the degree of indeterminacy (i.e.
neither membership, nor non-membership) as independent component in
1995 and de…ned the neutrosophic set. In 2003 W. B. Vasantha Kan-
dasamy and Florentin Smarandache introduced for the …rst time the I-
neutrosophic algebraic structures (such as neutrosophic semigroup, neutro-
sophic ring, neutrosophic vector space, etc.) based on neutrosophic num-
bers of the form a + bI, where ‘ I ’is the literal indeterminacy such that
I 2 = I, while a; b are real (or complex) numbers. In 2013 Smarandache
introduced the re…ned neutrosophic set, and in 2015 the re…ned neutro-
sophic algebraic structures built on sets on re…ned neutrosophic numbers
of the form a + b1I1 + b2I2 + : : : + bnIn, where I1; I2; : : : ; In are types of
sub-indeterminacies; in the same year he also introduced the (t; i; f)-
neutrosophic structures.
In 1971, Rosenfeld [53] …rst applied fuzzy sets to the study of algebraic
structures, and he initiated a novel notion called fuzzy groups. This pio-
neer work started a burst of studies on various fuzzy algebras. Kuroki [28]
studied fuzzy bi-ideals in semigroups and he examined some fundamental
properties of fuzzy semigroups in [28]. Mordesen [37] has demonstrated a
theoretical exposition of fuzzy semigroups and their application in fuzzy
coding, fuzzy …nite state machines and fuzzy languages. It is worth noting
that these fuzzy structures may give rise to more useful models in some
practical applications. The role of fuzzy theory in automata and formal

languages has extensively been discussed by Mordesen [37].
Pu and Liu [49] initiated the concept of fuzzy points and they also pro-
posed some inspiring ideas such as belongingness to (denoted by 2) and
quasi-coincidence (denoted by q) of a fuzzy point with a fuzzy set. Murali
[42] proposed the concept of belongingness of a fuzzy point to a fuzzy subset
under a natural equivalence on fuzzy subsets. These ideas played a vital role
to generate various types of fuzzy subsets and fuzzy algebraic structures.
Bhakat and Das [1, 2] applied these notions to introducing ( ; )-fuzzy
subgroups, where ; 2 f2; q; 2 _q; 2 ^qg and 6=2 ^q. Among ( ; )-
fuzzy subgroups, it should be noted that the concept of (2; 2_q)-fuzzy
subgroups is of vital importance since it is the most viable generalization
of the conventional fuzzy subgroups in Rosenfeld’s sense. Then it is natural
to investigate similar types of generalizations of the existing fuzzy sub-
systems of other algebraic structures. In fact, many authors have studied
(2; 2 _q)-fuzzy algebraic structures in di¤erent contexts [19, 22, 55]. Re-
cently, Shabir et al. [55] introduced (2; 2 _qk )-fuzzy ideals (quasi-ideals
and bi-ideals) of semigroups and gave various characterizations of particu-
lar classes of semigroups in terms of these fuzzy ideals. M. Khan introduced
the concept of (2 ; 2 _q )-fuzzy ideals and (2 ; 2 _q )-fuzzy soft ideals
in AG-groupoids
An AG-groupoid is an algebraic structure that lies in between a groupoid
and a commutative semigroup. It has many characteristics similar to that
of a commutative semigroup. If we consider x2 y 2 = y 2 x2 , which holds for all
x; y in a commutative semigroup, on the other hand one can easily see that
it holds in an AG-groupoid with left identity e and in AG -groupoids. In
addition to this xy = (yx)e holds for any subset fx, yg of an AG-groupoid.
This simply gives that how an AG-groupoid has closed connections with
commutative algebras.
We extend now for the …rst time the AG-Groupoid to the Neutrosophic
AG-Groupoid. A neutrosophic AG-groupoid is a neutrosophic algebraic
structure that lies between a neutrosophic groupoid and a neutrosophic
commutative semigroup.
Let M be an AG-groupoid under the law “:” One has (ab)c = (cb)a for
all a, b, c in M . Then M U I = fa + bI, where a; b are in M , and I is literal
indeterminacy such that I 2 = Ig is called a neutrosophic AG-groupoid. A
neutrosophic AG-groupoid in general is not an AG-groupoid.
If on MUI one de…nes the operation “ ”as: (a + bI) (c + dI) = ac + bdI,
then the neutrosophic AG-groupoid (M U I; ) is also an AG-groupoid since:
[(a1 + b1 I) (a2 + b2 I)] (a3 + b3 I) = [a1 a2 + b1 b2 I] (a3 + b3 I)

= (a1 a2 )a3 + (b1 b2 )b3 I
= (a3 a2 )a1 + (b3 b2 )b1 I:
Also
[(a3 + b3 I) (a2 + b2 I)] (a1 + b1 I) = [a3 a2 + b3 b2 I] (a1 + b1 I)

= (a3 a2 )a1 + (b3 b2 )b1 I:
In chapter one we discuss congruences in an AG-groupoid. In this chapter

we discuss idempotent separating congruence de…ned as: a b if and only
if (a 1 e)a = (b 1 e)b, in an inverse AG -groupoid S. We characterize in
two ways and show (a) that S= ' E; (E is the set of all idempotents of S)
if and only if E is contained in the centre of S, also it is shown; (b) that is
identical congruence on S if and only if E is self-centralizing. We show that
the relations min and max show are smallest and largest congruences on
S. Moreover we show that the relation de…ned as: a b if only if a 1 (ea) =
b 1 (eb), is a maximum idempotent separating congruence.
In chapter two we discuss gamma ideals in -AG -groupoid. Moreover
we show that a locally associative -AG -groupoid S has associative pow-
ers and S= , where a b implies that a bn = bn+1 , b an = an+1 8 a; b 2 S,
is a maximal separative homomorphic image of S. The relation is the
least left zero semilattice congruence on S, where is de…ne on S as a b
if and only if there exists some positive integers m, n such that bm 2 a S
and an 2 b S.
In chapter three we discuss embedding and direct products in AG-groupoids.
In chapter four we introduce the concept of left, right, bi, quasi, prime
(quasi-prime) semiprime (quasi-semiprime) ideals in AG-groupoids. We in-
troduce m system in AG-groupoids. We characterize quasi-prime and quasi-
semiprime ideals and …nd their links with m systems. We characterize
ideals in intra-regular AG-groupoids. Then we characterize intra-regular
AG-groupoids using the properties of these ideals.
In chapter …ve we introduce a new class of AG-groupoids namely strongly
regular and characterize it using its ideals.
In chapter six we introduce the fuzzy ideals in AG-groupoids and discuss
their related properties.
In chapter seven we characterize intra-regular AG-groupoids by the prop-
erties of the lower part of (2; 2 _q)-fuzzy bi-ideals. Moreover we character-
ize AG-groupoids using (2; 2 _qk )-fuzzy.
In chapter eight we discuss interval valued fuzzy ideals of AG-groupoids.
In chapter nine we characterize a Abel-Grassmann’s groupoid in terms
of its (2 ; 2 _q )-fuzzy ideals.
In chapter ten we characterize intra-regular AG-groupoids in terms of
(2 ; 2 _q )-fuzzy soft ideals.
1
Congruences on Inverse
AG-groupoids
In this chapter we discuss idempotent separating congruence de…ned as:
a b if and only if (a 1 e)a = (b 1 e)b, in an inverse AG -groupoid S. We
characterize in two ways and show (a) that S= ' E; (E is the set of
all idempotents of S) if and only if E is contained in the centre of S, also
it is shown; (b) that is identical congruence on S if and only if E is
self-centralizing. We show that the relations min and max are smallest
and largest congruences on S. Also we show that the relation de…ned
as: a b if only if a 1 (ea) = b 1 (eb), is a maximum idempotent separating
congruence.
1.1 AG-groupoids
The idea of generalization of a commutative semigroup was …rst introduced
by Kazim and Naseeruddin in 1972 (see [24]). They named it as a left almost
semigroup (LA-semigroup). It is also called an Abel-Grassmann’s groupoid
(AG-groupoid) [47].
An AG-groupoid is a non-associative and non-commutative algebraic
structure mid way between a groupoid and a commutative semigroup.
This structure is closely related with a commutative semigroup, because
if an AG-groupoid contains a right identity, then it becomes a commuta-
tive semigroup [43]. The connection of a commutative inverse semigroup
with an AG-groupoid has been given in [39] as: a commutative inverse semi-
group (S, ) becomes an AG-groupoid (S, ) under a b = b a 1 ; for all
a; b 2 S. An AG-groupoid (S; :) with left identity becomes a semigroup (S,
) de…ned as: for all x, y 2 S, there exists a 2 S such that x y = (xa)y
[47].
An AG-groupoid is a groupoid S whose elements satisfy the left invertive
law (ab)c = (cb)a, for all a; b; c 2 S: In an AG-groupoid, the medial law [24]
(ab)(cd) = (ac)(bd) holds for all a; b; c; d 2 S. An AG-groupoid may or may
not contains a left identity. If an AG-groupoid contains a left identity, then
it is unique [43]. In an AG-groupoid S with left identity, the paramedial law
(ab)(cd) = (db)(ca) holds for all a; b; c; d 2 S: If an AG-groupoid contains
a left identity, then it satis…es the following law
a(bc) = b(ac); for all a; b; c 2 S: (1)

Note that a commutative AG-groupoid S with left identity becomes a

commutative semigroup because if a; b and c 2 S. Then using left invertive
law and commutative law, we get
(ab)c = (cb)a = a(cb) = a(bc).
In [15] J. M. Howie de…ned a relation as (a; b) 2 if and only if

a 1 ea = b 1 eb on an inverse semigroup and show it maximum idempotent
separating congruence and characterize it in two ways. Also it is shown
that S= ' E if and only if E is central in S and that = 1S ; the identical
congruence on S; if and only if E is self centralizing in S: Moreover, J. M.
Howie in [14] de…ned a relations min and max as a min b if and only if
aa 1 bb 1 and 9e 2 E such that e aa 1 ea = eb and a max b if and only
if a 1 ea b 1 eb for all e 2 E and shown these as the smallest and largest
congruences on an inverse semigroup with trace . In this chapter, we
de…ned these congruences for inverse AG -groupoid and also characterize
it. An AG-groupoid S is called an inverse AG-groupoid if for every a 2 S
there exists a0 2 S such that (aa0 )a = a, (a0 a)a0 = a0 where a0 is an inverse
for a. We will write a 1 instead of a0 . If S is an inverse AG-groupoid, then
1
(ab) = a 1 b 1 and (a 1 ) 1 = a for all a; b 2 S.
Example 1 Let S = f1; 2; 3g and the binary operation “ ” de…ned on S as

follows:
1 2 3
1 2 2 3
2 2 2 2
3 2 2 2
Clearly S is non-associative and non-commutative because 2 = (1 1) 3 6=
1 (1 3) = 3 and 1 3 6= 3 1. (S; ) is an AG -groupoid without left identity.
Lemma 2 Let S be an inverse AG -groupoid and de…ned by a b, if

and only if aa 1 = bb 1 ; is a congruence relation.
Proof. Clearly is re‡exive, symmetric and transitive, so is an equiva-

lence relation. Let a b which implies that aa 1 = bb 1 , then we get.
1
(ac)(ac) = (ac)(a 1 c 1 ) = (aa 1 )(cc 1 ) = (bb 1
)(cc 1
)
= (bc)(b 1 c 1 ) = (bc)(bc) 1 :
1 1
Similarly we can show that (ca)(ca) = (cb)(cb) :
Lemma 3 Let S be an inverse AG -groupoid, then the relation = f(a; b) 2

S S : a 1 a = b 1 bg is a congruence on S:
Proof. It is available in [47].

Lemma 4 The congruence relation is equivalent to :

1 1
Proof. Let a b, this implies that a a=b b: Then we have
1
aa = ((aa 1 )a)a 1 = (a 1 a)(aa 1 ) = (b 1
b)(aa 1
)
= (a 1 a)(bb 1 ) = (b 1 b)(bb 1 )
= ((bb 1 )b)b 1 = bb 1 :
Thus a b:
1 1
Conversely, If a b; then aa = bb : Then
1
a a = a 1 ((aa 1 )a) = (aa 1 )(a 1 a) = (bb 1
)(a 1
a)
= (aa 1 )(b 1 b) = (bb 1 )(b 1 b)
= ((b 1 b)b 1 )b = b 1 b:
Hence a b:
Corollary 5 If is congruence on an inverse AG -groupoid, then (a; b) 2
, if and only if (a 1 ; b 1 ) 2 :
Proof. It is same as in [15].
Example 6 Let S = f1; 2; 3; 4g and the binary operation “ ” de…ned on S
as follows:
1 2 3 4
1 4 1 2 3
2 3 4 1 2
3 2 3 4 1
4 1 2 3 4
Clearly (S; ) is non-associative, non-commutative and it is an AG -groupoid
with left identity 4. Every element is an inverse of itself and so a 1 a =
aa 1 , for all a in S.
The following lemma is available in [47].
Lemma 7 The set E of all idempotents in an AG -groupoid forms a semi-
lattice structure.
1.2 Inverse AG -groupoids

In the rest, by S we shall mean an inverse AG -groupoid in which aa 1 =
a 1 a; holds for every a 2 S:
Let be a congruence on S. The restriction of to E, is congruence on
E, which we call trace of and is denoted by = tr : The set ker = fa 2
S=(9e 2 E)a eg is the kernel of :
Theorem 8 Let E be the set of all idempotents of S and let be a congru-

ence on E, then the relation min = f(a; b) 2 S S : aa 1 bb 1 and there
exist e 2 E; e aa 1 and ea = ebg is the smallest congruence on S with
trace :
Proof. Clearly is re‡exive. Now let a min b, this implies that aa 1 bb 1
and there exist e 2 E such that e aa 1 and ea = eb: As e aa 1 and
aa 1 bb 1 which implies that e bb 1 also eb = ea which implies that
b min a, which shows that min is symmetric. Again let a min b and b min c
which implies that aa 1 bb 1 cc 1 this implies that aa 1 cc 1 . Also
e aa 1 and f bb 1 for e; f 2 E. Since is compatible so, ef (aa 1 )(aa 1 ) =
aa 1 which implies that ef aa 1 : Now ea = eb implies thatf (ea) = f (eb)
so we have
f (ea) = (f f )(ea) = (ae)(f f ) = (ae)f = (f e)a; and
f (eb) = (f f )(eb) = (be)(f f ) = (be)f = (f e)b
Also f b = f c implies that e(f b) = e(f c). Now
e(f b) = (ee)(f b) = (bf )(ee) = (bf )e = (ef )b = (f e)b
e(f c) = (ee)(f c) = (cf )(ee) = (cf )e = (ef )c = (f e)c
Hence (f e)a = (f e)c which shows that min is transitive.
Now let a min b, then
1
(ca)(ca) = (ca)(c 1 a 1 ) = (cc 1 )(aa 1 ) (cc 1
)(bb 1
)
= (cb)(c 1 b 1 ) = (cb)(cb) 1 , and
1 1 1
(cc )e (cc )(aa ) = (ca)(c 1 a 1 )
= (ca)(ca) 1 , where (cc 1
)e 2 E, and
1 1
((cc )e)(ca) = ((cc )c)(ea)
1 1
= ((cc )c)(eb) = ((cc )e)(cb):
Therefore ca min ca:
1 1 1
Again let a min b then by de…nition aa bb ; e aa and ea = eb
Now
1
(ac)(ac) = (ac)(a 1 c 1 )
= (aa 1 )(cc 1 ) (bb 1 )(cc 1 ) = (bc)(bc) 1
and
e(cc 1 ) (aa 1 )(cc 1 ) = (ac)(a 1 c 1 )
= (ac)(ac) 1 , where e(cc 1 ) 2 E
Also
1 1 1 1
(e(cc ))(ac) = (ea)((cc )c) = (eb)((cc )c) = (e(cc ))(bc):
Thus ac min bc. Therefore min is a congruence relation.

The remaining proof is same as in [14].
Theorem 9 Let E be the set of all idempotents of S and let be a con-
gruence on E, then the relation max = f(a; b) 2 S S : (8e 2 E)
a 1 (ea) b 1 (eb)g is the largest congruence on S with trace :
Proof. Clearly max is an equivalence relation as is an equivalence rela-
tion on E.
Let us suppose that a max b; then a 1 (ea) b 1 (eb) so
1
(ac) (e(ac)) = (a 1 c 1 ) ((ee)(ac)) = (a 1 c 1 )((ea)(ec))
= (a 1 (ea))(c 1 (ea)) (b 1 (eb))(c 1 (ec))
= (b 1 c 1 )((eb)(ec)) = (bc) 1 (e(bc)):
Thus ac max bc: Similarly ca max cb: Therefore max is congruence on S.
Remaining proof is same as in [14].
The relation 1s = f(x; x) : x 2 Sg is a congruence relation which we
call the identical congruence. A congruence whose trace is the identical
congruence 1 is called idempotent separating.
Theorem 10 Let E be the set of all idempotents of S and let the relation
de…ned as a b if and only if (a 1 e)a = (b 1 e)b; for any e in E, is an
idempotent separating congruence on S:
Proof. It is easy to prove that is an equivalence relation. Now let a b;
then (a 1 e)a = (b 1 e)b; for every idempotent e in E; now we get
1
((ac) e)(ac) = ((a 1 c 1 )(ee))(ac) = ((a 1 e)(c 1 e))(ac)
= ((a 1 e)a)((c 1 e)c) = ((b 1 e)b)(c 1 e)c
= ((b 1 e)(c 1 e))(bc) = ((bc) 1 e)(bc):
Thus ac bc: Similarly ca cb: Hence is a congruence relation on S:
Now let e f for e; f in E: Then for every g in E; (e 1 g)e = (f 1 g)f
so by (1), we have eg = f g: The equality holds in particular when g = e.
Hence e = f g. Similarly for g = f; we obtain ef = f: Since ef = f e; so
e = f: Thus is idempotent separating.
If E is the semilattice of an inverse semigroup S; we de…ne E , the
centralizer of E in S, by
E = fz 2 S : ez = ze for every e in Eg:

Clearly E E If E = S, then the idempotents are central. If E = E,
we shall say that E is self-centralizing.
Theorem 11 Let E be the set of all idempotents of S and let be the
idempotent separating congruence on S: Then Ker = E where E be the
centralizer of E in S.
Proof. Let S be an inverse AG -groupoid and let be the idempotent

separating congruence on S: Let a 2 Ker ; so a f for some f 2 E: also
a 1 f 1 = f; so a 1 a f; implies that a aa 1 . So for all e in E (a 1 e)a =
((a 1 a) 1 e)(a 1 a), then we get
1 1 1 1 1
((a a) e)(a a) = ((aa )e)(a a)
1 1 1
= (e(a a))(a a) = (a a)e; that is
1 1
(a e)a = (a a)e: (5)
Also we have,
ea = e((aa 1 )a) = (aa 1 )(ea) = ((ea)a 1 )a

= ((a 1 a)e)a = ((a 1 e)a)a = (aa)(a 1 e)
= (ea 1 )(aa) = ((aa)a 1 )e = ((aa 1 )a)e = ae:
Thus a 2 E :
Conversely, assume that a 2 E : Then for all e in E; ae = ea; so
1
(a e)a = (ae)a 1 = (ea)a 1 = (ea)((a 1 a)a 1 )
= (a 1 a)((ea)a 1 ) = (a 1 a)((a 1 a)e)
= (e(a 1 a))(aa 1 ) = ((aa 1 ) 1 e)(aa 1 ):
1
Thus a aa and so a 2 Ker . Hence E = Ker .
idempotent separating congruence on S: Then (a; b) 2 if and only if
a 1 a = b 1 b; and ab 1 2 E . Dually (a; b) 2 if and only if aa 1 = bb 1
and a 1 b 2 E :
Proof. Let (a; b) 2 , then (a 1 e)a = (b 1 e)b which implies that (ae)a 1
=
(be)b 1 for all e in E:
Now ((a 1 e)a)((ae)a 1 ) = ((b 1 e)b)((be)b 1 ) which implies that
1 1 1 1
((a a)e)(aa ) = ((b b)e)(bb ): (6)
Therefore we get
1
a a = ((a 1 a)a 1 )((aa 1 )a) = ((a 1 a)(aa 1 ))(a 1 a)
= ((a 1 a)(aa 1 ))(aa 1 ) = ((b 1 b)(aa 1 ))(bb 1 )
= ((aa 1 )(b 1 b))(bb 1 ) = ((bb 1 )(b 1 b))(aa 1 )
= ((b 1 b)(b 1 b))(aa 1 ) = (b 1 b)(aa 1 ) = (a 1 a)(b 1
b):
Similarly we can show that b 1 b = (a 1 a)(b 1 b). Therefore a 1 a = b 1 b:

Now let (a; b) 2 , then (a 1 e)a = (b 1 e)b which implies that (ae)a 1 =
(be)b 1 , which implies that (a((ae)a 1 ))b 1 = (a((be)b 1 ))b 1 .
Now we obtain
1 1
(a((ae)a ))b = ((ae)(aa 1 ))b 1 = (b 1 (aa 1 ))(ae)
= (a(b 1 a 1 ))(ae) = (ea)((b 1 a 1 )a)
= (ea)((aa 1 )b 1 ) = (e(aa 1 ))(ab 1 ); and
1 1
(a((be)b ))b = ((be)(ab 1 ))b 1 = (b 1 (ab 1 ))(be)
= (b 1 b)((ab 1 )e) = (aa 1 )((ab 1 )e)
= (ab 1 )((aa 1 )e) = (ab 1 )(e(aa 1 )):
Hence ab 1 2 E :
Conversely, let a 1 a = b 1 b and ab 1 2 E , then e(ab 1 ) = (ab 1 )e for
all e 2 E, which implies that (a 1 (e(ab 1 )))b = (a 1 ((ab 1 )e))b: Now we
get
1 1
(a (e(ab )))b = (b(e(ab 1 )))a 1 = (e(b(ab 1 )))a 1
= (e(a(bb 1 )))a 1 = ((ee)(a(aa 1 )))a 1
= (((aa 1 )a)(ee))a 1 = (ae)a 1 = (a 1 e)a;
Now
1 1
(a ((ab )e))b = (a 1 ((ab 1 )(ee)))b = (a 1 ((ae)(b 1 e)))b
= ((ae)(a 1 (b 1 e)))b = (((b 1 e)a 1 )(ea))b
= (((b 1 e)e)(a 1 a))b = ((eb 1 )(a 1 a))b
= ((eb 1 )(b 1 b))b = (b(b 1 b))(eb 1 )
= (b 1 e)((b 1 b)b) = (b 1 e)((bb 1 )b)
= (b 1 e)b:
Therefore (a 1 e)a = (b 1 e)b. Hence a b:
Let a b then by de…nition a 1 e a = b 1 e b: Now as a 1 a = b 1 b so
aa 1 = bb 1 :
Now as a 1 e a = b 1 e b which implies that (a 1 (a 1 e a))b =
(a 1 ( b 1 e b))b:
So we get
1 1 1 1 1 1
(a (a e a))b = a e a a b= a e b b b
1 1 1 1
= b b b a e = ea bb b
1 1 1
= ea b = ba e = ba (ee)
1 1
= (ee) a b =e a b ; and
Now we get
1 1 1 1 1 1
(a ( b e b))b = (b( b e b))a =( b e (bb))a
1 1 1 1
= ( b b (eb))a = (e (bb b))a
1 1
= (eb)a = (a b)e:
1
Hence a b2E :

idempotent separating congruence on S: Then S= ' E if and only if E is
central in S:
Proof. Since is idempotent separating congruence so S= is a semilattice

if each -class contains atmost one idempotent. Thus if S= is semilattice
then S= = E: Let us suppose that each class contains an idempotent
that is for every x 2 S, there exist an f 2 E such that f x which implies
that f f 1 = xx 1 and f 1 x 2 E , thus
1 1 1
x = (xx )x = (f f )x = f x2E ;
but this holds for any x in S; so E = S:

1
Conversely, suppose that E = S, then xf 2 S = E and
1 1 1 1 1 1 1
xx = (xx )(xx ) = (xx )(xx ) = ff ;
1
Then by theorem 5, x f , that is, x xx ; which shows that every class
contains an idempotent.

idempotent separating congruence on S: Then = 1S ; the identical con-
gruence on S; if and only if E is self centralizing in S:
Proof. Let = 1S ; Then for z 2 E implies that ze = ez; for all e 2 E if

we write f for zz 1 then zz 1 = f = f f = f f 1 also we get
1 1 1 1
zf e = ef z = (ef ) z = z(ef ) = z ef = e zf :
Therefore zf 1 2 E : Then by theorem 5, z zz 1 ; but = 1S ; so z =

zz 1 2 E: Thus E = E:
Conversely, assume that E = E: Let x y then x 1 x = y 1 y and
xy 1 2 E = E; since xy 1 is idempotent so (xy 1 ) 1 = xy 1 , implies
that x 1 y = xy 1 , also (x 1 ; y 1 ) 2 so
1
xx = ((xx 1 )x)x 1 = ((yy 1 )x)x 1
= ((xy 1 )y)x 1 = ((x 1 y)y)x 1
= (x 1 y)(x 1 y) = x 1 y:
Also we get
x = (xx 1 )x = (yy 1
)x = (xy 1
)y
= (x 1 y)y = (xx 1
)y = (yy 1
)y = y:
Hence = 1S :
Theorem 15 Let E be the set of all idempotents of S then the relation

de…ned by a b if only if a 1 (ea) = b 1 (eb) is a maximum idempotent sep-
arating congruence on S:
Proof. Clearly is an equivalence relation. Let a b; which implies that
a 1 (ea) = b 1 (eb). Now
1
(ac) (e(ac)) = (a 1 c 1 )(e(ac)) = (a 1 c 1 )((ee)(ac))
= (a 1 c 1 )((ea)(ec)) = (a 1 (ea))(c 1 (ec))
= (b 1 (eb))(c 1 (ec)) = (b 1 c 1 )((eb)(ec))
= (bc) 1 (e(bc)):
Therefore ac bc: Similarly ca cb: Hence is a congruence relation. Now
suppose that e f , where e, f 2 E; then for every idempotent g we have
e 1 (ge) = f 1 (gf ); which implies that ge = gf: In particular when g = e,
then ee = ef , implies that e = ef and for g = f; f e = f f implies that
f e = f; but since ef = f e implies that e = f: Thus is idempotent
separating congruence. Now let be any other idempotent separating con-
gruence. We shall show that : Let (x; y) 2 then (x 1 ; y 1 ) 2 ;
since is congruence, it follows that x y which implies that ex ey, also
x 1 (ex) y 1 (ey), but both x 1 (ex) and y 1 (ey) are idempotents, and so
it follows that x 1 (ex) = y 1 (ey): Thus x y: Hence is maximum.
Theorem 16 Let E be the set of all idempotents of S then the relation
de…ned on S with = f(a; b) 2 S S (8e 2 E) : ((a 1 )2 e)a2 = ((b 1 )2 e)b2 g
is a congruence relation on S:
Proof. It is clear that is an equivalence relation. Now suppose that a b
and c is an arbitrary element of S; then
1 2 2 2
(((ac) ) e) (ac) = ((a 1 c 1 )2 e) (ac)
= ( (a 1 )2 (c 1 )2 e)(a2 c2 )
= (((a 1 )2 e)(c 1 )2 )(a2 c2 )
= (((a 1 )2 e)a2 )((c 1 )2 c2 )
= (((b 1 )2 e)b2 )((c 1 )2 c2 )
= (((b 1 )2 e)(c 1 )2 )(b2 c2 )
= ((b 1 )2 (c 1 )2 e)(b2 c2 )
1 1 2 2
= ((b c ) e) (bc)
1 2 2
= (((bc) ) e) (bc) :
Thus (ac; bc) 2 : Similarly (ca; cb) 2 : Hence is congruence relation.
Lemma 17 Let E be the set of all idempotents of S then the centralizer

E of E in S; is an inverse subgroupoid of S.
Proof. Let a; b 2 E ; then ae = ea and be = eb; for all e 2 E, so
(ab) e = (ab) (ee) = (ae) (be) = (ea) (eb) = (ee) (ab) = e (ab) :
Therefore E is a subgroupoid of S.
1 1 1
Now let a 2 E then ae = ea implies that (ae) = (ea) or a e=
1 1
ea ; so a 2 E : Hence E is an inverse subgroupoid.
Theorem 18 Let S be an inverse AG -groupoid with semilattice E and
let be the maximum idempotent separating congruence on S then S= is
fundamental.
Proof. Every idempotent in S= has the form e . Let us suppose that
1 1
(a ; b ) 2 S= then for every e in E (a ) ((e ) (a )) = (b ) ((e ) (b ))
which implies that a 1 (ea) = b 1 (eb) ; consequentlya 1 (ea) b 1 (eb)
but is idempotent separating so a 1 (ea) = b 1 (eb) that is a b implies
that a = b so S= is identical. Thus S= is fundamental.
2
Structural Properties of
-AG -groupoids
In this chapter we discuss gamma ideals in -AG -groupoids. We show
that a locally associative -AG -groupoid S has associative powers and
S= , where a b implies that a bn = bn+1 , b an = an+1 8 a; b 2 S, is a
maximal separative homomorphic image of S. The relation is the least
left zero semilattice congruence on S, where is de…ne on S as a b if
and only if there exists some positive integers m, n such that bm 2 a S
and an 2 b S.
2.1 Gamma Ideals in -AG-groupoids

Let S and be any non-empty sets. If there exists a mapping S S!S
written as (x; ; y) by x y; then S is called a -AG-groupoid if x y 2 S
such that the following -left invertive law holds for all x; y; z 2 S and
; 2
(x y) z = (z y) x: (1)
A -AG-groupoid also satis…es the -medial law for all w; x; y; z 2 S and
; ; 2
(w x) (y z) = (w y) (x z): (2)
Note that if a -AG-groupoid contains a left identity, then it becomes
an AG-groupoid with left identity.
A -AG-groupoid is called a -AG -groupoid if it satis…es the following
law for all x; y; z 2 S and ; 2
x (y z) = y (x z): (3)
A -AG -groupoid also satis…es the -paramedial law for all w; x; y; z 2

S and ; ; 2
(w x) (y z) = (z y) (x w): (4)
De…nition 19 Let S be a -AG-groupoid, a non-empty subset A of S is
called -AG-subgroupoid if a b 2 A for all a, b 2 A and 2 or A is
called -AG-subgroupoid if A A A:
De…nition 20 A subset A of a -AG-groupoid S is called -left (right)
ideal of S if S A A (A S A) and A is called -two-sided ideal of S if
it is both -left and -right ideal.
De…nition 21 A -AG-subgroupoid A of a -AG-groupoid S is called a

-bi-ideal of S if (A S) A A.
-interior ideal of S if (S A) S A:
De…nition 23 A -AG-groupoid A of a -AG-groupoid S is called a -
quasi-ideal of S if S A \ A S A:
-(1; 2)-ideal of S if (A S) (A A) A.
De…nition 25 A -two-sided ideal P of a -AG-groupoid S is called -
prime ( -semiprime) if for any -two-sided ideals A and B of S, A B
P (A A P ) implies either A P or B P (A P ) :
De…nition 26 An element a of an -AG-groupoid S is called an intra-
regular if there exist x; y 2 S and ; ; 2 such that a = (x (a a)) y
and S is called an intra-regular -AG-groupoid S, if every element of S is
an intra-regular.
Example 27 Let S = f1; 2; 3; 4; 5; 6; 7; 8; 9g. The following multiplication
table shows that S is an AG-groupoid and also an AG-band.
1 2 3 4 5 6 7 8 9
1 1 4 7 3 6 8 2 9 5
2 9 2 5 7 1 4 8 6 3
3 6 8 3 5 9 2 4 1 7
4 5 9 2 4 7 1 6 3 8
5 3 6 8 2 5 9 1 7 4
6 7 1 4 8 3 6 9 5 2
7 8 3 6 9 2 5 7 4 1
8 2 5 9 1 4 7 3 8 6
9 4 7 1 6 8 3 5 2 9
It is easy to observe that S is a simple AG-groupoid that is there is no
left or right ideal of S. Now let = f , , g de…ned as follows.
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
1 2 2 2 2 2 2 2 2 2 1 8 8 8 8 8 8 8 8 8
2 2 2 2 2 2 2 2 2 2 2 8 8 8 8 8 8 8 8 8
3 2 2 2 2 2 2 2 2 2 3 8 8 8 8 8 8 8 8 8
4 2 2 2 2 2 2 2 2 2 4 8 8 8 8 8 8 8 8 8
5 2 2 2 2 2 2 2 2 2 5 8 8 8 8 8 8 8 8 8
6 2 2 2 2 2 2 2 2 2 6 8 8 8 8 8 8 8 8 8
7 2 2 2 2 2 2 2 2 2 7 8 8 8 8 8 8 8 8 8
8 2 2 2 2 2 2 2 2 2 8 8 8 8 8 8 8 8 8 8
9 2 2 2 2 2 2 2 2 2 9 8 8 8 8 8 8 8 8 8
1 2 3 4 5 6 7 8 9
1 8 8 8 8 8 8 8 8 8
2 8 8 8 8 8 8 8 8 8
3 8 8 8 8 8 8 8 8 8
4 8 8 8 8 8 8 8 8 8
5 8 8 8 8 8 8 8 8 8
6 8 8 8 8 8 8 8 8 8
7 8 8 8 8 8 8 8 8 8
8 8 8 8 8 8 8 8 8 8
9 8 8 8 8 8 8 8 9 9
It is easy to prove that S is a -AG-groupoid because (a b) c = (c b) a for
all a, b, c 2 S and , 2 . Clearly S is non-commutative and non-
associative because 8 9 6= 9 8 and (1 2) 3 6= 1 (2 3).
Example 28 Let S = f1; 2; 3; g. The following Cayley’s table shows that
S is an AG-groupoid.
1 2 3
1 2 3 1
2 1 2 3
3 3 1 2
Let us de…ne = f , , g as follows.
1 2 3 1 2 3 1 2 3
1 1 1 1 1 2 2 2 1 1 1 1
2 1 1 1 2 2 2 2 2 1 1 1
3 1 1 1 3 2 2 3 3 1 1 3
Clearly S is an intra-regular -AG-groupoid because 1 = (2 (1 1)) 3;

2 = (1 (2 2)) 3; 3 = (3 (3 3)) 3:
Theorem 29 A -AG -groupoid S is an intra-regular -AG -groupoid
if S a = S or a S = S holds for all a 2 S.
Proof. Let S be a -AG -groupoid such that S a = S holds for all a 2 S;
then S = S S. Let a 2 S and therefore, by using (2); we have
a 2 S = (S S) S = ((S a) (S a)) S = ((S S) (a a)) S

= (S (a a)) S:
Which shows that S is an intra-regular -AG -groupoid.

Let a 2 S and assume that a S = S holds for all a 2 S; then by using
(1), we have
a 2 S = S S = (a S) S = (S S) a = S a:
Thus S a = S holds for all a 2 S and therefore it follows from above that
S is an intra-regular.
Corollary 30 If S is a -AG -groupoid such that a S = S holds for all
a 2 S; then S a = S holds for all a 2 S:
Theorem 31 If S is an intra-regular -AG -groupoid, then (B S) B =
B \ S; where B is a -bi-( -generalized bi-) ideal of S.
Proof. Let S be an intra-regular -AG -groupoid, then clearly (B S) B
B \ S. Now let b 2 B \ S which implies that b 2 B and b 2 S; then since
S is an intra-regular -AG -groupoid so there exist x; y 2 S and ; ; 2
such that b = (x (b b)) y: Now we have
b = (x (b b)) y = (b (x b)) y = (y (x b)) b

= (y (x ((x (b b)) y))) b = (y ((x (b b)) (x y))) b
= ((x (b b)) (y (x y))) b = (((x y) y) ((b b) x)) b
= ((b b) (((x y) y) x)) b = ((b b) ((x y) (x y))) b
= ((b b) ((x x) (y y))) b = (((y y) (x x)) (b b)) b
= (b (((y y) (x x)) b)) b 2 (B S) B:
Which shows that (B S) B = B \ S:

Corollary 32 If S is an intra-regular -AG -groupoid, then (B S) B =
B; where B is a -bi-( -generalized bi-) ideal of S.
Theorem 33 If S is an intra-regular -AG -groupoid, then (S I) S =
S \ I; where I is a -interior ideal of S.
Proof. Let S be an intra-regular -AG -groupoid, then clearly (S I) S
S \ I. Now let i 2 S \ I which implies that i 2 S and i 2 I; then since S
is an intra- regular -AG -groupoid so there exist x; y 2 S and ; ; 2
such that i = (x (i i)) y: Now we have
i = (x (i i)) y = (i (x i)) y = (y (x i)) i

= (y (x i)) ((x (i i)) y) = (((x (i i)) y) (x i)) y
= ((i x) (y (x (i i)))) y = (((y (x (i i))) x) i) y 2 (S I) S:
Which shows that (S I) S = S \ I:

Corollary 34 If S is an intra-regular -AG -groupoid, then (S I) S =
I; where I is a -interior ideal of S.
Lemma 35 If S is an intra-regular regular -AG -groupoid, then S =
S S:
Proof. It is simple.
Lemma 36 A subset A of an intra-regular -AG -groupoid S is a -left

ideal if and only if it is a -right ideal of S:
Proof. Let S be an intra-regular -AG -groupoid and let A be a -right

ideal of S; then A S A: Let a 2 A and since S is an intra-regular -AG -
groupoid so there exist x; y 2 S and ; ; 2 such that a = (x (a a)) y:
Let p 2 S A and 2 ; then by we have
p = s a = s ((x (a a)) y) = (x (a a)) (s y) = (a (x a)) (s y)

= ((s y) (x a)) a = ((a x) (y s)) a = (((y s) x) a) a
= (a a) ((y s) x) = (x (y s)) (a a) = a ((x (y s)) a) 2 A S A:
Which shows that A is a -left ideal of S:

Let A be a -left ideal of S; then S A A: Let a 2 A and since S is an
intra-regular -AG -groupoid so there exist x; y 2 S and ; ; 2 such
that a = (x (a a)) y: Let p 2 A S and 2 ; then we have
p = a s = ((x (a a) y) s = (s y) (x (a a)) = ((a a) x) (y s)

= ((y s) x) (a a) = (a a) (x (y s)) = ((x (y s)) a) a 2 S A A:
Which shows that A is a -right ideal of S:
Theorem 37 In an intra-regular -AG -groupoid S, the following condi-

tions are equivalent.
(i) A is a -bi-( -generalized bi-) ideal of S.

(ii) (A S) A = A and A A = A:
Proof. (i) =) (ii) : Let A be a -bi-ideal of an intra-regular -AG -
groupoid S; then (A S) A A. Let a 2 A, then since S is an intra-regular
so there exist x; y 2 S and ; ; 2 such that a = (x (a a)) y: Now we
have
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x ((x (a a)) y))) a = (y ((x (a a)) (x y))) a
= ((x (a a)) (y (x y))) a = ((a (x a)) (y (x y))) a
= ((a y) ((x a) (x y))) a = ((x a) ((a y) (x y))) a

= ((x a) ((a x) (y y))) a = (((y y) (a x)) (a x)) a
= (a (((y y) (a x)) x)) a 2 (A S) A:
Thus (A S) A = A holds. Now we have
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x ((x (a a)) y))) a = (y ((x (a a)) (x y))) a
= ((x (a a)) (y (x y))) a = ((a (x a)) (y (x y))) a
= (((y (x y)) (x a)) a) a = (((a x) ((x y) y)) a) a
= (((a x) ((y y) x)) a) a = (((a (y y)) (x x)) a) a
= ((((x x) (y y)) a) a) a
= ((((x x) (y y)) ((x (a a)) y)) a) a
= ((((x x) (y y)) ((a (x a)) y)) a) a
= ((((x x) (a (x a))) ((y y) y)) a) a
= (((a ((x x) (x a))) ((y y) y)) a) a
= (((a ((a x) (x x))) ((y y) y)) a) a
= ((((a x) (a (x x))) ((y y) y)) a) a
= ((((a a) (x (x x))) ((y y) y)) a) a
= (((((y y) y) (x (x x))) (a a)) a) a
= ((a ((((y y) y) (x (x x))) a)) a) a
((A S) A) A A A:
Hence A = A A holds.
(ii) =) (i) is obvious.
(i) A is a -(1; 2)-ideal of S.
(ii) (A S) (A A) = A and A A = A:
Proof. (i) =) (ii) : Let A be a -(1; 2)-ideal of an intra-regular -AG -
groupoid S; then (A S)(A A) A and A A A. Let a 2 A, then since
S is an intra- regular so there exist x; y 2 S and ; ; 2 such that
a = (x (a a) y: Now
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x ((x (a a)) y))) a = (y ((x (a a)) (x y))) a
= ((x (a a)) (y (x y))) a = (((x y) y) ((a a) x)) a
= (((y y) x) ((a a) x)) a = ((a a) (((y y) x) x)) a
= ((a a) ((x x) (y y))) a = (a ((x x) (y y))) (a a) 2 (A S) (A A):
Thus (A S) (A A) = A: Now we have
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x a)) ((x (a a)) y)
= (x (a a)) ((y (x a)) y)

= (a (x a)) ((y (x a)) y)
= (((y (x a)) y) (x a)) a
= ((a x) (y (y (x a)))) a
= ((((x (a a)) y) x) (y (y (x a)))) a
= (((x y) (x (a a))) (y (y (x a)))) a
= (((x y) y) ((x (a a)) (y (x a)))) a
= (((y y) x) ((x (a a)) (y (x a)))) a
= (((y y) x) ((x y) ((a a) (x a)))) a
= (((y y) x) ((a a) ((x y) (x a)))) a
= ((a a) (((y y) x) ((x y) (x a)))) a
= ((a a) (((y y) x) ((x x) (y a)))) a
= ((((x x) (y a)) ((y y) x)) (a a)) a
= ((((a y) (x x)) ((y y) x)) (a a)) a
= (((((x x) y) a) ((y y) x)) (a a)) a
= (((x (y y)) (a ((x x) y))) (a a)) a
= ((a ((x (y y)) ((x x) y))) (a a)) a
= ((a ((x (x x)) ((y y) y))) (a a)) a
2 ((A S) (A A)) A A A:
Hence A A = A:
(i) A is a -interior ideal of S.
(ii) (S A) S = A:
Proof. (i) =) (ii) : Let A be a -interior ideal of an intra-regular -AG -
groupoid S; then (S A) S A. Let a 2 A, then since S is an intra- regular
so there exist x; y 2 S and ; ; 2 such that a = (x (a a)) y: Now we
have
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x a)) ((x (a a)) y)
= (((x (a a)) y) (x a)) y
= ((a x) (y (x (a a)))) y
= (((y (x (a a))) x) a) y 2 (S A) S:
Thus (S A) S = A:

(i) A is a -quasi ideal of S.

(ii) S Q \ Q S = Q:
Proof. (i) =) (ii) : Let Q be a -quasi ideal of an intra-regular -AG -
groupoid S; then S Q \ Q S Q. Let q 2 Q, then since S is an intra-
regular so there exist x, y 2 S and ; ; 2 such that q = (x (q q)) y:
Let p q 2 S Q; for some 2 ; then
p q = p ((x (q q)) y) = (x (q q)) (p y) = (q (x q)) (p y)

= (q p) ((x q) y) = (x q) ((q p) y) = (y (q p)) (q x)
= q ((y (q p)) x) 2 Q S:
Now let q y 2 Q S; then we have
q p = ((x (q q)) y) p = (p y) (x (q q))

= x ((p y) (q q)) = x ((q q) (y p))
= (q q) (x (y p)) = ((x (y p)) q) q 2 S Q:
Hence Q S = S Q: Then we have
q = (x (q q)) y = (q (x q)) y = (y (x q)) q 2 S Q:
Thus q 2 S Q \ Q S implies that S Q \ Q S = Q.



(ii) A is a -two-sided two-sided ideal of S:
Proof. (i) =) (ii) : Let S be an intra-regular -AG -groupoid and let
A be a -(1; 2)-ideal of S; then (A S) (A A) A: Let a 2 A, then since
S is an intra-regular so there exist x; y 2 S and ; ; 2 ; such that
a = (x (a a)) y: Now let 2 ; then
s a = s ((x (a a)) y) = (x (a a)) (s y)

= (a (x a)) (s y) = ((s y) (x a)) a
= ((s y) (x a)) ((x (a a)) y)
= (x (a a)) (((s y) (x a)) y)
= (y ((s y) (x a))) ((a a) x)
= (a a) ((y ((s y) (x a))) x)
= (x (y ((s y) (x a)))) (a a)
= (x (y ((a x) (y s)))) (a a)
= (x ((a x) (y (y s)))) (a a)
= ((a x) (x (y (y s)))) (a a)
= ((((x (a a)) y) x) (x (y (y s)))) (a a)
= (((x y) (x (a a))) (x (y (y s)))) (a a)
= ((((a a) x) (y x)) (x (y (y s)))) (a a)
= ((((y x) x) (a a)) (x (y (y s)))) (a a)
= (((y (y s)) x) ((a a) ((y x) x))) (a a)
= (((y (y s)) x) ((a a) ((x x) y))) (a a)
= ((a a) (((y (y s)) x) ((x x) y))) (a a)

= ((((x x) y) ((y (y s)) x)) (a a)) (a a)
= (a (((x x) y) (((y (y s)) x) a))) (a a) 2 (A S) (A A) A:
Hence A is a -left ideal of S and so A is a -two-sided ideal of S:

(ii) =) (i) : Let A be a -two-sided ideal of S. Let y 2 (A S) (A A) ;
then y = (a s) (b b) for some a; b 2 A, s 2 S and ; ; 2 : Now we have
y = (a s) (b b) = b ((a s) b) 2 A S A:
Hence (A S) (A A) A and therefore A is a -(1; 2)-ideal of S.


(ii) A is a -interior ideal of S:
Proof. (i) =) (ii) : Let A be a -(1; 2)-ideal of an intra-regular -AG -
0
groupoid S; then (A S) (A A) A: Let p 2 (S A) S; then p = (s a) s
0
for some a 2 A, s; s 2 S and ; 2 . Since S is intra-regular so there
exist x; y 2 S and ; ; 2 such that a = (x (a a)) y: Now we have

0 0
p = (s a) s = (s ((x (a a)) y)) s
0 0
= ((x (a a)) (s y)) s = (s (s y)) (x (a a))
0 0
= (s (s y)) (a (x a)) = a ((s (s y)) (x a))
0
= ((x (a a)) y) ((s (s y)) (x a))
0
= ((a (x a)) y) ((s (s y)) (x a))
0
= ((a (x a)) (s (s y))) (y (x a))
0
= ((a s ) ((x a) (s y))) (y (x a))
0
= ((a s ) ((y s) (a x))) (y (x a))
0
= ((a s ) (a ((y s) x))) (y (x a))
0
= ((a a) (s ((y s) x))) (y (x a))
0
= ((a a) ((y s) (s x))) (y (x a))
0
= ((y (x a)) ((y s) (s x))) (a a)
0
= ((y (y s)) ((x a) (s x))) (a a)
0
= ((y (y s)) ((x s ) (a x))) (a a)
0
= ((y (y s)) (a ((x s ) x))) (a a)
0
= (a ((y (y s)) ((x s ) x))) (a a)
2 (A S) (A A) A:
Thus (S A) S A: Which shows that A is a -interior ideal of S:
(ii) =) (i) : Let A be a -interior ideal of S; then (S A) S A: Let
p 2 (A S) (A A); then p = (a s) (b b); for some a; b 2 A, s 2 S and
; ; 2 . Since S is intra-regular so there exist x; y 2 S and ; ; 2
such that a = (x (a a)) y: Now we have
p = (a s) (b b) = ((b b) s) a
= ((b b) s) ((x (a a)) y)
= (x (a a)) (((b b) s) y)
= ((((b b) s) y) (a a)) x
= ((a a) (y ((b b) s))) x
= (((y ((b b) s)) a) a) x 2 (S A) S A:
Thus (A S) (A A) A.
Now by using (3) and (4); we have
A A A S = A (S S) = S (A S) = (S S) (A S)
= (S A) (S S) = (S A) S A:
Which shows that A is a -(1; 2)-ideal of S:

(i) A is a -bi-ideal of S:
(ii) A is a -interior ideal of S:
Proof. (i) =) (ii) : Let A be a -bi-ideal of an intra-regular -AG -
0
groupoid S; then (A S) A A: Let p 2 (S A) S; then p = (s a) s for
0
some a 2 A, s; s 2 S and ; 2 . Since S is an intra-regular so there
exist x; y 2 S and ; ; 2 such that a = (x (a a)) y: Now we have
0 0
p = (s a) s = (s ((x (a a)) y)) s
0 0
= ((x (a a)) (s y)) s = (s (s y)) (x (a a))
0
= ((a a) x) ((s y) s )
0 0
= (((s y) s ) x) (a a) = ((x s ) (s y)) (a a)
0 0
= (a a) ((s y) (x s )) = (((s y) (x s )) a) a
0
= (((s y) (x s )) ((x (a a)) y)) a
0
= (((s y) (x (a a))) ((x s ) y)) a
0
= ((((a a) x) (y s)) ((x s ) y)) a
0
= ((((x s ) y) (y s)) ((a a) x)) a
0
= ((a a) ((((x s ) y) (y s)) x)) a
0
= ((x (((x s ) y) (y s))) (a a)) a
0
= (a ((x (((x s ) y) (y s))) a)) a
2 (A S) A A:
Thus (S A) S A: Which shows that A is a -interior ideal of S:

(ii) =) (i) : Let A be a -interior ideal of S; then (S A) S A: Let
p 2 (A S) A; then p = (a s) b for some a; b 2 A, s 2 S and ; 2 :
Since S is an intra-regular so there exist x; y 2 S and ; ; 2 such that
b = (x (b b)) y: Now
p = (a s) b = (a s) ((x (b b)) y) = (x (b b)) ((a s) y)

= (((a s) y) (b b)) x = ((b b) (y (a s))) x
= (((y (a s)) b) b) x 2 (S A) S A:
Thus (A S) A A:
Now
A A A S = A (S S) = S (A S) = (S S) (A S)
= (S A) (S S) = (S A) S A:
Which shows that A is a -bi-ideal of S:



(ii) A is a -quasi ideal of S:
Proof. (i) =) (ii) : Let A be a -(1; 2)-ideal of intra-regular -AG -
groupoid S; then (A S) (A A) A: Now we have
S A = S (A A) = S ((A A) A)
= (A A) (S A) = (A S) (A A) A:
and by using (1) and (3); we have
A S = (A A) S = ((A A) A) S = (S A) (A A) = (S (A A)) (A A)
= ((S S) (A A)) (A A) = ((A A) (S S)) (A A)
= (A S) (A A) A:
Hence (A S) \ (S A) A: Which shows that A is a -quasi ideal of S:

(ii) =) (i) : Let A be a -quasi ideal of S; then (A S) \ (S A) A:
Now A A A S and A A S A: Thus A A (A S) \ (S A) A:
Then
(A S) (A A) = (A A) (S A) A (S A) = S (A A) S A:
and
(A S) (A A) = (A A) (S A) A (S A) = S (A A)
= (S S) (A A) = (A A) (S S) A S:
Thus (A S) (A A) (A S) \ (S A) A: Which shows that A is a

-(1; 2)-ideal of S.
Lemma 45 Let A be a subset of an intra-regular -AG -groupoid S, then
A is a -two-sided ideal of S if and only if A S = A and S A = A.
Theorem 46 For an intra-regular -AG -groupoid S the following state-
ments are equivalent.
(i) A is a -left two-sided ideal of S.

(ii) A is a -right two-sided ideal of S.
(iii) A is a -two-sided ideal of S.
(iv) A S = A and S A = A.
(v) A is a -quasi ideal of S.
(vi) A is a -(1; 2)-ideal of S.
(vii) A is a -generalized bi-ideal of S.
(viii) A is a -bi-ideal of S.
(ix) A is a -interior ideal of S.
Proof. (i) =) (ii) and (ii) =) (iii) are easy.
(iii) =) (iv) is followed by above Lemma and (iv) =) (v) is obvious.
(v) =) (vi) It is easy.
(vi) =) (vii) : Let A be a -(1; 2)-ideal of an intra-regular -AG -
groupoid S, then (A S) (A A) A. Let p 2 (A S) A; then p = (a s) b
for some a; b 2 A, s 2 S and ; 2 . Now since S is an intra-regular so
there exist x; y 2 S and ; ; 2 such that such that b = (x (b b)) y
then we have
p = (a s) b = (a s) ((x (b b)) y)
= (x (b b)) ((a s) y) = (y (a s)) ((b b) x)
= (b b) ((y (a s)) x) = (x (y (a s))) (b b)
= (x (a (y s))) (b b)
= (a (x (y s))) (b b) 2 (A S) (A A) A:
Which shows that A is a -generalized bi-ideal of S.

(vii) =) (viii) is simple.
(viii) =) (ix) is followed easily.
(ix) =) (i) is followed by previous results .
Theorem 47 In a -AG -groupoid S, the following conditions are equiv-
alent.
(i) S is intra-regular.
(ii) Every -bi-ideal of S is -idempotent.
Proof. (i) =) (ii) is obvious.
(ii) =) (i) : Since S a is a -bi-ideal of S, and by assumption S a is
-idempotent, so we have
a 2 (S a) (S a) = ((S a) (S a)) (S a)
= ((S S) (a a)) (S a) (S (a a)) (S S)
= (S (a a)) S.
Hence S is intra-regular.
Lemma 48 If I and J are -two-sided ideals of an intra-regular -AG -
groupoid S , then I \ J is a -two-sided ideal of S:
Lemma 49 In an intra-regular -AG -groupoid I J = I \ J, for every
-two-sided ideals I and J in S.
Proof. Let I and J be any -two-sided ideals of S, then obviously I J
I \ J. Since I \ J I and I \ J J, then (I \ J) (I \ J) I J, also,
I \ J is a -two-sided ideal of S; so we have I \ J = (I \ J) (I \ J) I J.

Hence I J = I \ J.
Lemma 50 Let S be a -AG -groupoid, then S is an intra-regular if and
only if every -left ideal of S is -idempotent.
Proof. Let S be an intra-regular -AG -groupoid, then every -two-sided
ideal of S is -idempotent.
Conversely, assume that every -left ideal of S is -idempotent. Since
S a is a -left ideal of S, so we have
a 2 S a = (S a) (S a) = ((S a) (S a)) (S a)
= ((S S) (a a)) (S a) (S (a a)) (S S)
= (S (a a)) S.
Lemma 51 In an AG -groupoid S, the following conditions are equiva-
lent.
(ii) A = (S A)(S A); where A is any -left ideal of S.
Proof. (i) =) (ii) : Let A be a -left ideal of an intra-regular -AG -
groupoid S; then S A A and then, (S A)(S A) = S A A: Now
A = A A S A = (S A)(S A); which implies that A = (S A)(S A):
(ii) =) (i) : Let A be a -left ideal of S; then A = (S A)(S A) A A;
which implies that A is -idempotent and so S is an intra-regular.
Theorem 52 A -AG -groupoid S is called -totally ordered under in-
clusion if P and Q are any -two-sided ideals of S such that either P Q
or Q P .
A -two-sided ideal P of a -AG -groupoid S is called -strongly irre-
ducible if A \ B P implies either A P or B P , for all -two-sided
ideals A, B and P of S.
Lemma 53 Every -two-sided ideal of an intra-regular -AG -groupoid
S is -prime if and only if it is -strongly irreducible.
Proof. It is an easy.
Theorem 54 Every -two-sided ideal of an intra-regular -AG -groupoid
S is -prime if and only if S is -totally ordered under inclusion.
Proof. Assume that every -two-sided ideal of S is -prime. Let P and
Q be any -two-sided ideals of S, so , P Q = P \ Q, and P \ Q is a
-two-sided ideal of S, so is prime, therefore P Q P \ Q; which implies
that P P \ Q or Q P \ Q; which implies that P Q or Q P . Hence
S is -totally ordered under inclusion.
Conversely, assume that S is -totally ordered under inclusion. Let I, J

and P be any -two-sided ideals of S such that I J P . Now without
loss of generality assume that I J then
I=I I I J P.
Therefore either I P or J P , which implies that P is -prime.

Theorem 55 The set of all -two-sided ideals of an intra-regular -AG -
groupoid S, forms a -semilattice structure.
Proof. Assume that I be the set of all -two-sided ideals of an intra-
regular -AG -groupoid S and let A, B 2 I , since A and B are -two-
sided ideals of S, then by using (2); we have
(A B) S = (A B) (S S) = (A S) (B S) A B.
Also S (A B) = (S S) (A B) = (S A) (S B) A B.
Thus A B is a -two-sided ideal of S. Hence I is closed. Also we have,

A B = A \ B = B \ A = B A, which implies that I is commutative, so
is associative. Now A A = A, for all A 2 I . Hence I is -semilattice.
Theorem 56 For an intra-regular -AG -groupoid S, the following state-
ments holds.
(i) Every -right ideal of S is -semiprime.
(ii) Every -left ideal of S is -semiprime.
(iii) Every -two-sided ideal of S is -semiprime
Proof. (i) : Let R be a -right ideal of an intra-regular -AG -groupoid
S. Let a a 2 R for some 2 and let a 2 S: Now since S is an intra-
regular so there exist x; y 2 S and ; ; 2 such that a = (x (a a)) y.
Now we have
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x a)) ((x (a a)) y) = (x (a a)) ((y (x a)) y)
= (x (y (x a))) ((a a) y)
= (a a) ((x (y (x a))) y) 2 R (S S) = R S R:
Which shows that R is -semiprime.

(ii) : Let L be a -left ideal of S: Let a a 2 L for some 2 and let
a 2 S now since S is an intra-regular so there exist x; y 2 S and ; ; 2
such that a = (x (a a)) y; then we have
a = (x (a a)) y = (a (x a)) y = (y (x a)) a

= (y (x a)) ((x (a a)) y) = (x (a a)) ((y (x a)) y)
= (y (y (x a))) ((a a) x) = (a a) ((y (y (x a))) x)
= (x (y (y (x a)))) (a a) 2 S L L:
Which shows that L is -semiprime.

(iii) is obvious.
Theorem 57 A -two-sided ideal of an intra-regular -AG -groupoid S

is minimal if and only if it is the intersection of two minimal -two-sided
ideals.
Proof. Let S be an intra-regular -AG -groupoid and Q be a minimal
-two-sided ideal of S, let a 2 Q. As S (S a) S a and S (a S)
a (S S) = a S; which shows that S a and a S are -left ideals of S so
S a and a S are -two-sided ideals of S.
Now
S (S a \ a S) \ (S a \ a S) S
= S (S a) \ S (a S) \ (S a) S \ (a S) S
(S a \ a S) \ (S a) S \ S a S a \ a S:
Which implies that S a \ a S is a -quasi ideal so S a \ a S is a

-two-sided ideal.
Also since a 2 Q, we have
S a\a S S Q\Q S Q\Q Q.
Now since Q is minimal so S a \ a S = Q; where S a and a S are

minimal -two-sided ideals of S, because let I be a -two-sided ideal of S
such that I S a; then
I \a S S a\a S Q;
which implies that

I \ a S = Q: Thus Q I:
So we have
S a S Q S I I; gives
S a = I:
Thus S a is a minimal -two-sided ideal of S. Similarly a S is a mini-

mal -two-sided ideal of S:
Conversely, let Q = I \ J be a -two-sided ideal of S, where I and J
are minimal -two-sided ideals of S; then Q is a -quasi ideal of S, that is
S Q \ Q S Q:
0 0
Let Q be a -two-sided ideal of S such that Q Q, then
0 0 0
S Q \Q S S Q\Q S Q; also S Q S I I
0
and Q S J S J.
Now
0 0 0
S S Q = (S S) S Q = Q S (S S)
0 0 0
= Q S S = (S S) Q = S Q
0
implies that S Q is a -left ideal and hence a -two-sided ideal. Similarly
0
Q S is a -two-sided ideal of S.
But since I and J are minimal -two-sided ideals of S, so
0 0
S Q = I and Q S = J:
But Q = I \ J; which implies that,
0 0 0
Q=S Q \Q S Q:
0
Which give us Q = Q . Hence Q is minimal.
2.2 Locally Associative -AG -groupoids

In this section we introduce a new non-associative algebraic structure namely
locally associative -AG -groupoids and decompose it using -congruences.
An AG-groupoid S is called a locally associative -AG-groupoid if (a a) a =
a (a a), holds for all a in S and ; 2 . If S is a locally associa-
tive AG-groupoid then it is easy to see that (S a) S = S (a S) or
(S S) S = S (S S). For particular 2 , let us denote a a = a2
2
for some 2 and a a = a , 8 2 i.e. a a = a2 and generally
n
a a a:::a a = a (n times.)
Let S be an -AG -groupoid and a relation be de…ned on S as follows:
a b if and only if there exists a positive integer n such that a bn = bn+1
and b an = an+1 , for all a and b in S:
Proposition 58 If S is a locally associative -AG -groupoid, then a an+1 =

(an+1 ) a, for all a in S and positive integer n.
Proof.
a an+1 = a (an a) = an (a a) = (an 1 a) (a a) = (a a) (a an 1

)
= (a a) an = (an a) a = (an+1 ) a:
Proposition 59 In a locally associative -AG -groupoid S; am an = am+n

8 a 2 S and positive integers m, n.
Proof.
am+1 an = (am a) an = (an a) am = (a an ) am

= (am an ) a = am+n a = am+n+1 :
Proposition 60 If S is a locally associative -AG -groupoid, then for all

a, b in S; (a b)n = an bn and positive integer n 1 and (a b)n = bn an ,
for n 2.
Proof.
(a b)2 = (a b) (a b) = (a a) (b b) = a2 b2 :
(a b)k+1 = (a b)k (a b) = (ak bk ) (a b) = (ak a) (bk b) = ak+1 bk+1 .
Let n 2. Then by (3) and (1), we get
(a b)n = an bn = (a an 1 ) (b bn 1 ) = b ((a an 1 ) bn 1 ))
= b ((bn 1 an 1 ) a) = b ((b a)n 1 a) = (b a)n 1 (b a)
= (b a)n = bn an .
Proposition 61 In a locally associative -AG -groupoid S, (am )n = amn

for all a 2 S and positive integers m, n.
Proof.
n(m+1)
(am+1 )n = (am a)n = (am )n an = amn an = amn+n = a :
Theorem 62 Let S be a locally associative -AG -groupoid. If a bm =

bm+1 and b an = an+1 for a; b 2 S and positive integers m, n, then a b.
Proof. If n > m, then
bn m (a bm ) = bn m bm+1
a (bn m bm ) = bn m+m+1
a bn m+m = bn+1
a bn = bn+1 .
Theorem 63 The relation on a locally associative -AG -groupoid is

a congruence relation.
Proof. Evidently is re‡exive and symmetric. For transitivity we may

proceed as follows.
Let a b and b c so that there exist positive integers n, m such that
a bn = bn+1 , b an = an+1 , and
b cm = cm+1 , c bm = bm+1 .
Let k = (n + 1)(m + 1) 1, that is, k = n(m + 1) + m. Thus we get,
n(m+1)+m n(m+1)
a ck = a c = a (c cm ) = a f(cm+1 )n cm g
m(n+1)
= a f(b cm )n cm g = a f(bn cmn ) cm g = a (c bn )
m(n+1) m(n+1) m(n+1)
= c (a bn ) = c bn+1 = (cm b)n+1 = bn+1 c
= (b cm )n+1 = ck+1 .
Similarly, c ak = ak+1 . Thus is an equivalence relation. To show that

is compatible, assume that a b such that for some positive integer n,
a bn = bn+1 and b an = an+1 .
Let c 2 S, then, we get
(a c) (b c)n = (a c) (bn cn ) = (a bn ) (c cn ) = bn+1 cn+1 = (b c)n+1 .
Similarly, (b c) (a c)n = (a c)n+1 . Hence is a congruence relation on

S.
Lemma 64 Let S be a locally associative -AG -groupoid, then a b b a; for
all a; b in S.
Proof.
(a b) (b a)n+1 = (a b) (an+1 bn+1 ) = (a an+1 ) (b bn+1 )

= an+2 bn+2 = (b a)n+2 .
Similarly, (b a) (a b)n+1 = (a b)n+2 . Hence a b b a, for all a; b in S.
A relation on an AG-groupoid S is called separative if a b a2 and

a b b2 implies that a b.
Theorem 65 The relation is separative.
Proof. Let a, b 2 S, a b a2 , and a b b2 . Then by de…nition of there
exist positive integers m and n such that,
(a b) (a2 )m = (a2 )m+1 , a2 (a b)m = (a b)m+1 and

(a b) (b2 )n = (b2 )n+1 , b2 (a b)n = (a b)n+1 .
Then
(a b) a2m = (a b) (am am ) = (a am ) (b am ) = (am+1 ) (b am )

= b (am+1 am ) = b a2m+1 , but (a b) a2m = (a2 )m+1 = a2m+2 ,
which implies that b a2m+1 = a2m+2 . Also (a b) (b2 )n = (b2 )n+1 , implies
that b2n+1 a = b2n+2 . Also, we get
b2n+2 b2 = (b2n+1 a) b2 ,
this implies that
b2n+4 = b2 (a b2n+1 ) = a (b2 b2n+1 ) = a b2n+3 .
Hence, a b.
Theorem 66 Let S be a locally associative -AG -groupoid. Then S

is a maximal separative commutative image of S.
Proof. is separative, and hence S is separative. We now show that

is contained in every separative congruence relation on S. Let a b
so that there exists a positive integer n such that,
a bn = bn+1 and b an = an+1 .
We need to show that a b, where is a separative congruence on S. Let

k be any positive integer such that,
a bk bk+1 and b ak ak+1 . (5)
Suppose k > 3.
(a bk 1 2
) = (a bk 1
) (a bk 1
) = a2 b2k 2
= (a a) (bk 2
bk )
= (a bk 2
) (a bk ) = (a bk 2
) bk+1
Therefore
(a bk 2
) (a bk ) (a bk 2
) bk+1 .
Thus we get
(a bk 2
) bk+1 = (bk+1 bk 2
) a = b2k 1
a = (bk bk 1
) a = (a bk 1
) bk .
k 1 k 1 2k 1 k 1
Also (a b ) bk = (bk b ) a=b a = (b bk ) a
k 1
= (a bk ) b ,
implies that
(a bk 1 2
) (a bk ) bk 1
.
Since a bk bk+1 and (a bk ) bk 1

bk+1 bk 1
, hence (a bk 1 2
) (bk )2 .
It further implies that,
(a bk 1 2
) (a bk 1
)bk (bk )2 .
Thus
a bk 1
bk .
Similarly,
b ak 1
ak .
Thus if (13) holds for k, it holds for k 1.
Now obviously (13) yields
a b3 0
b4 and b a3 0
a4 .
Also, we get
(a b3 ) a2 0 b4 a2 and (b a3 ) b2 0 a4 b2
(a2 b3 ) a 0 b4 a2 and (b2 a3 ) b 0 a4 b2
(b3 a2 ) a 0 a2 b4 and (a3 b2 ) b 0 b2 a4
a3 b3 0 a2 b4 and b3 a3 0 b2 a4
a3 b3 0 a2 b4 and a3 b3 0 b2 a4 ;
which implies that (b2 a)2 0 a3 b3 0 (a2 b)2 ; and as 0 is separative

and (b2 a) (a2 b) = (b2 a2 ) (a b) = (a2 b2 ) (a b) = a3 b3 , so a2 b 0 b2 a.
Now we get
(a2 b) a 0 (b2 a) a
(a b) a2 0 a2 b2
a2 (b a) 0 a2 b2
b a3 0 a2 b2 but b a3 0
a4 ;
Thus (b a)2 0 b a3 0 (a2 )2 , now since 0 is separative and a2 (b a) =

b a3 , so we get b a 0 a2 .
Similarly we can obtain a b 0 b2 .
Also it is easy to show that (13) holds for k = 2.
Thus if (5) holds for k, it holds for k = 1. By induction down from k,
it follows that (5) holds for k = 1, a b b2 and b a a2 . Now it is easy
to see that a b b2 , we get (b a)2 b3 a, and again from a b b2 we
get b3 a b4 . So (b a)2 b3 a b4 implies that b a b2 which further
implies that a b b a. Thus we obtain a b. Hence and so S
is the maximal separative commutative image of S.
Lemma 67 If x a = x (a = a2 ) for some x in a locally associative -
AG -groupoid then xn a = xn for some positive integer n.
Proof. Let n = 2, then using (2), we get

x2 a = (x x) (a a) = (x a) (x a) = x x = x2 .
Let the result be true for k, that is, xk a = xk . Then by (2) and Proposition
1, we get
xk+1 a = (x xk ) (a a) = (x a) (xk a) = x xk = xk+1 .
Hence xn a = xn for all positive integers n.
Lemma 68 If S is a -AG-groupoid, then Q = fx j x 2 S, x a = x and
a = a2 g is a commutative subsemigroup.
Proof. As a a = a, we have a 2 Q. Now if x, y 2 Q, then by identity (2),
x y = (x a) (y a) = (x y) (a a) = (x y) a.
To prove that Q is commutative and associative, assume that x, y and z
belong to Q. Then by using (1), we get
x y = (x a) y = (y a) x = y x. Also
(x y) z = (z y) x = x (y z).
Hence Q is a commutative subsemigroup of S.
Theorem 69 Let and be separative congruences on locally asso-
ciative -AG -groupoid S and x2 a = x2 (a = a2 ) for all x in S. If
\ (Q Q ) \ (Q Q ), then .
Proof. If x y then,
(x2 (x y))2 (x2 (x y) (x2 y 2 ) (x2 y 2 )2 .
It follows that (x2 (x y))2 , (x2 y 2 )2 2 Q . Now by (2), (1), (3); respec-
tively, we get,
(x2 (x y)) (x2 y 2 ) = (x2 x2 ) ((x y) y 2 ) = (x2 x2 ) (y 3 x)
= x4 (y 3 x) = y 3 (x4 x) = y 3 x5 and
(y 3 x5 ) a = (y 3 x5 ) (a a) = (y 3 a) (x5 a) = y 3 x5 .
So x2 (x y) (x2 y 2 ) 2 Q. Hence (x2 (x y))2 (x2 (x y) (x2 y 2 ) (x2 y 2 )2
implies that
x2 (x y) x2 y 2 :
Since x2 y 2 x4 and (x2 y 2 ), x4 2 Q. Thus x2 y 2 x4 and we get
(x2 )2 x2 (x y) (x y)2 which implies that x2 x y. Finally, x2 y 2
and x2 , y 2 2 Q, implying that x2 y 2 , x2 x y y 2 . Thus x y because
is separative.
Lemma 70 Every left zero congruence is commutative.

Proof. Let a a and b b which implies that a b a b, (a b) (a b) (a b)2 =
(b a)2 and so we obtain a b b a.
The relation de…ne on S by a b if and only if there exists some
positive integers m, n such that bm 2 a S and an 2 b S.
Theorem 71 Let S be a locally associative -AG -groupoid. Then the
relation is the least semilattice congruence on S.
Proof. The relation is obviously re‡exive and symmetric. To show tran-
sitivity, let a b and b c, where a; b; c 2 S. Then a x = bm for some x
and b y = cn , for some x and y 2 S. Then we get
cmn = (cn )m = (b y)m = y m bm = y m (a x) = a (y m x);
implies that ck = a z, where k = mn and z = (y m x). Similarly, b x0 =

0 0
am0 and c y 0 = bn implies that ak = c z 0 .
Let a; b; c 2 S and a b , (9 m; n 2 Z + )(9 x; y 2 S) bm = a x; an =
b y. If m = 1; n > 1, that is b = a x, an = b y for some x; y 2 S, then
b3 = (b b) (a x) = a (b2 x) 2 a S.
Similarly we can consider the case m = n = 1. Suppose that m; n > 1.

Then we obtain
(b c)m = bm cm = (a x) cm = (a x) (c cm 1 )
= (a c) (x cm 1 ) = (a c) y, where y = x cm 1
:
Thus a c b c and c a c b.
Now to show that is a semilattice congruence on S, …rst we need to
show that a b implies a b a.
Let a b, then bm = a x and an = b y for some x and y 2 S. So
(a b)m = am bm = am (a x) = a (am x).
Also an = b y implies that an+2 = a2 an = (a a) (b y) = (a b) (a y).

Hence a b a which implies that a2 a, (a2 ) = (a ) and so S is
idempotent.
Next we show that is commutative. By Proposition 4, (a b)2 =
2
(b a) , which shows that a b b a that is a b = b a , that is S is
a commutative AG-groupoid and so is left zero commutative semigroup of
idempotents. Therefore is a semilattice congruence on S. Next we will
show that is contained in any other left zero semilattice congruence
on S. Let a b, then bm = a x and an = b y. Now since a a2 and b b2 ,
it implies that a x a2 x, a an and b bm which further implies that
a b y and b a x. It is easy fact that a b b a, for some 2 . Also
since b b2 and is compatable, so we get b y b2 y. We can easily see

that b a a b a b y b2 y which implies that b a b2 y. Similarly
we can show that a b a2 x. So a b y b2 yb a a b a2 x a x b
implies that a b. Thus is a least semilattice congruence on S.
Theorem 72 is separative.
Proof. Let a2 a b and a b b2 , then there exist positive integers m, m
and n ; n such that:
(a2 )m = (a b)2 x; (a b)m = (a2 )2 x and

(a b)n = (b2 )2 y ; (b2 )n = (a b)2 y.
Now we get,
a2m+2 = a2m a2 = (a2 )m a2 = ((a b)2 x) a2

= (a2 x) (a b)2 = (a2 x) (a2 b2 ) = (a2 x) (b2 a2 )
= b2 ((a2 x) a2 ) = b2 t6 , where t6 = ((a2 x) a2 ).
Similarly,
b2n+2 = b2n b2 = ((a b)2 y) b2 = (b2 y) (a2 b2 ) = a2 ((b2 y) b2 )

= a2 t7 , where t7 = ((b2 y) b2 ).
Hence is separative.
Theorem 73 Let S be a locally associative -AG -groupoid. Then S
is a maximal semilattice separative image of S.
Proof. By Theorem 6, is the least semilattice congruence on S and
S is a semilattice. Hence S is a maximal semilattice separative
image of S.
2.3 Decomposition to Archimedean Locally

Associative AG-subgroupoids
Theorem 74 Every locally associative -AG -groupoid S is uniquely ex-
pressible as a semilattice Y of Archimedean locally associative -AG -
groupoids (S ) ( 2 Y ). The semilattice Y is isomorphic with the max-
imal semilattics separative image S of S and (S ) ( 2 Y ) are the
equivalence classes of S mod .
Proof. is least semilattice congruence on S. Next we will prove that
equivalence classes mod are Archimedean locally associative -AG -
groupoids and the semilattice Y is isomorphic to S . Let a; b 2 (S ) ,
where 2 Y , then a b implies that am 2 b S, bn 2 a S, so am = b x

and bn = a y; where x; y 2 S. If x 2 S , 6= then = , then
we get am+1 = a am = a (b x) = b (a x) 2 b (S ) = b (S ) .
Similarly one can show that bn+1 2 a (S ) . This shows that (S ) is right
Archimedean and so is locally associative Archimedean -AG -groupoid
S. Next we show the uniqueness. Let S be a semilattice Y of Archimedean
AG -groupoid (S ) , 2 Y . We need to show that (S ) are equivalence
classes of S mod . Let a; b 2 S:Then we show that a b if and only if a
and b belong to the same (S ) . If a and b both belong to the same (S ) ,
then each divides the power of the other. Since (S ) is Archimedean,
a b by de…nition. Conversely, if a b then a x = bm and b y = an for
some x; y 2 S and some m; n 2 Z + . If x 2 (S@ ) , then a x 2 (S @ )
and bm 2 (S ) , so that @ = . Hence , in the semilattice Y . By
symmetry, it follows that that is = .
3
Embedding and Direct Product
of AG-groupoids
3.1 Embedding in AG-groupoids
In this chapter we prove that under certain conditions a right cancellative
AG -groupoid can be embedded in a cancellative commutative monoid
whose special type of elements form an abelian group and the identity of
this group coincides with the identity of the commutative monoid.
An element a in an AG-groupoid S is called left cancellative, if ab = ac
implies that b = c. Similarly, c is right cancellative, if ac = bc implies that
a = b.
In this chapter we shall consider that S is a right cancellative AG -
groupoid with left identity and T is a subgroupoid of S such that elements
of S commute with elements of T 2 . A relation has been introduced on
the subset N of S T 2 , so that we obtain an AG-groupoid with right
identity. We have proved that N= is a cancellative commutative monoid.
A mapping from S to N= has been de…ned to show that it is in fact an
epimorphism from S to a commutative sub-monoid A, of N= . At the end
it has been shown that special type of elements of N= form an Abelian
group.
2
Lemma 75 If S is an AG -groupoid, then (ab) = a2 b2 = b2 a2 , for all
a; b in S.
Proof. By (2) and (4), we get (ab)2 = (ab)(ab) = (aa) (bb) = a2 b2 , also
(ab)2 = (ab)(ab) = (ba)(ba) = b2 a2 .
Example 76 Let S = fa; b; cg; and the binary operation ( ) be de…ned on

S as follows:
a b c
a c a b
b b c a
c a b c
Then (S; ) is an AG -groupoid with left identity c. Clearly it is cancella-

tive.
Example 77 Let S = f1; 2; 3; 4g; the binary operation ( ) be de…ned on S

as follows:
1 2 3 4
1 1 2 3 4
2 4 3 3 3
3 3 3 3 3
4 2 3 3 3
It is non-commutative and non-associative because 4 = 1 4 6= 4 1 = 2;
2 = (2 1) 1 6= 2 (1 1) = 4. (S; ) is an AG -groupoid. The subset
A = f2; 4g, of S, is a commutative sub-semigroup of S.
3.2 Main Results

Theorem 78 If T is a subgroupoid of a right cancellative AG -groupoid
S with left identity and elements of S commute with elements of T 2 , then
S becomes a commutative monoid.
Proof. Let N = f(si t2j ; t2k ) : si 2 S and tj ; tk 2 T g, clearly N is closed be-

cause by (2) and lemma 75, we get (si t2j ; t2k )(sl t2m ; t2n ) = ((si sl )(tj tm )2 ; (tk tn )2 ),
for all si ; sl 2 S and tj ; tm ; tk ; tn 2 T . De…ne a relation on N as (si t2j ,
t2k ) (sl t2m , t2n ) if and only if (si t2j )t2n = (sl t2m )t2k . It is easy to prove that is
re‡exive and symmetric. To prove that is transitive, we proceed as follows.
Let (si t2j , t2k ) (sl t2m , t2n ) and (sl t2m , t2n ) sp t2q , t2r . Then (si t2j )t2n = (sl t2m )t2k
and (sl t2m )t2r = (sp t2q )t2n . Multiply the …rst equation from left by t2r , then
by lemma 75, we obtain t2n ((si t2j )t2r ) = t2n ((sp t2q )t2k ) which implies that
(si t2j )t2r = (sp t2q )t2k , thus (si t2j , t2k ) sp t2q , t2r , proving that is transitive.
If (si t2j , t2k ) (sl t2m , t2n ), then (si t2j )t2n = (sl t2m )t2k , now we get (t2n t2j )si =
(t2k t2m )sl . Multiplying this equation by sp from left side and we get (t2n t2j )(sp si ) =
(t2k t2m )(sp sl ), now multiply this equation by t2q t2r from right side and using
lemma 75, we get ((si t2j )(sp t2q ))(t2n t2r ) = (sl t2m )(sp t2q )(t2k t2r ). Thus
((si t2j )(sp t2q ); t2k t2r ) ((sl t2m )(sp t2q ); t2n t2r )
that is,
(si t2j ; t2k )(sp t2q ; t2r ) (sl t2m ; t2n )(sp t2q ; t2r ):
This shows that is right compatible. Similarly we can show that is
left compatible. Hence is a congruence relation on N .
Let M = N= = f[(si t2j ; t2k )] : si 2 S and tj ; tk 2 T g where [(si t2j ; t2k )]
represents any class in N= . Then it is easy to see that M is an AG -
groupoid. Clearly [(t2o ; t2o )] is the right identity in M , where t0 is an ar-
bitrary element of T , because if [(si t2j ; t2k )] is an arbitrary element in M ,
then ((si t2j )t2o )t2k = (si t2j )(t2k t2o ). Therefore ((si t2j )t2o ; t2k t2o ) (si t2j ; t2k ) which
implies that (si t2j ; t2k )(t2o ; t2o ) (si t2j ; t2k ) or [(si t2j ; t2k )][(t2o ; t2o )] = [(si t2j ; t2k )].
Hence [(t2o ; t2o )] is the right identity in M . Since M is an AG -groupoid

with right identity so it will become a commutative monoid.
Let tx be any …xed element of T . We de…ne a mapping : S ! M
by (si ) = [(si t2x ; t2x )], for all si 2 S and tx 2 T . Suppose si ; sj 2 S
such that si = sj . Then clearly [(si t2x ; t2x )] = [(sj t2x ; t2x )] for tx 2 T . Thus
(si ) = (sj ) . This shows that is well de…ned. Next we show that
(si sj ) = (si ) (sj ) . Since (si ) (sj ) = [((si sj )(t2x t2x ); t2x t2x )]. Also using
lemma 75, we get ((si sj )(t2x t2x ))t2x = (t2x (t2x t2x ))(si sj )) = ((t2x t2x )t2x )(si sj )) =
((si sj )t2x )(t2x t2x ), this implies that ((si sj )(t2x t2x ); t2x t2x ) ((si sj )t2x ; t2x ) and so
[((si sj )(t2x t2x ); t2x t2x )] = ((si sj )t2x ; t2x ) = (si sj ) . Hence (si ) (sj ) = (si sj ) .
This shows that is a homomorphism.
It is one-to-one, because (si ) = (sj ) implies that [(si t2x ; t2x )] = [(sj t2x ; t2x )],
that is, (si t2x ; t2x ) (sj t2x ; t2x ). Thus (si t2x )t2x = sj t2x t2x , which implies that
si = sj .
If A = f[(si t2x ; t2x )] : si 2 S and tx 2 T g. Then A M and monomor-
phism : S ! A is onto. As for every [(si t2x ; t2x )] in A there exists si such
that (si ) = [(si t2x ; t2x )]. Clearly [(t2o ; t2o )] belongs to A.
Lemma 79 A right cancellative AG-groupoid with left identity is left can-

cellative.
Proof. It is easy.
Since S contains the left identity so it is easy to see that [(t2j ; t2k )] 2 M .
Now we prove the following theorem.
Theorem 80 M is cancellative and elements of the form [(t2i ; t2j )] in M;

form an Abelian group.
Proof. Let us suppose that (si t2j ; t2k )(sp t2q ; t2r ) (sl t2m ; t2n )(sp t2q ; t2r ), that is,
[(si t2j ; t2k )][(sp t2q ; t2r )] = [(sl t2m ; t2n )][(sp t2q ; t2r )]
which implies that
[(si t2j )(sp t2q ); t2k t2r )] = [(sl t2m )(sp t2q ); t2n t2r )];
Then we get,
[(si sp )(t2j t2q ); t2k t2r )] = [(sl sp )(t2m t2q ); t2n t2r )]
which implies that
((si sp )(t2j t2q ))(t2n t2r ) = ((sl sp )(t2m t2q ))(t2k t2r ):
Now lemma 75, we get
((si sp )(t2n t2j ))(t2r t2q ) = ((sl sp )(t2k t2m ))(t2r t2q );
now since S is right cancellative so we get (si sp )(t2n t2j ) = (sl sp )(t2k t2m )
which by lemma 75 implies that sp ((t2n t2j )si ) = sp ((t2k t2m )sl ), therefore by
lemma 79, we get (t2n t2j )si = (t2k t2m )sl , using we get,(si t2j )t2n = (sl t2m )t2k .
Thus (si t2j ; t2k ) (sl t2m ; t2n ). Hence M is right cancellative. Similarly we can
show that M is left cancellative. Now using lemma 75, we can easily
see that (t2i t2j )t2o = (t2j t2i )t2o which implies that (t2i t2j ; t2o ) (t2j t2i ; t2o ), that is,
[(t2i ; t2j )][(t2j ; t2i )] = [(t2o ; t2o )]. Thus [(t2i ; t2j )] is the inverse of [(t2j ; t2i )]. Hence
all the cancellative elements [(t2i ; t2j )] of M form an Abelian group G in M .
We note that the product of two cancellative elements of G, is in G. We
have proved in theorem 1, that [(t2o ; t2o )] is the identity element of M , since
G contains elements of the form [(t2x ; t2y )], therefore [(t2o ; t2o )] is in G which
is unique since G is a group.
3.3 Direct Products in AG-groupoids

In this section we show that the direct product of regular AG-groupoids
is the most generalized class of the direct product of an AG-groupoids.
It has proved that the direct product of weakly regular, intra-regular,
right regular, left regular, left quasi regular, completely regular and (2; 2)-
regular AG-groupoids with left identity coincide. Also we have proved that
the direct product of intra-regular AG-groupoids with left identity (AG -
groupoids) is regular but the converse is not true in general. Further we
have shown that non-associative direct product of regular, weakly regu-
lar, intra-regular, right regular, left regular, left quasi regular, completely
regular, (2; 2)-regular and strongly regular AG -groupoids do not exist.
If S1 and S2 are AG-groupoids, then S1 S 2 = f(s1 ; s2 ) : s1 2 S1 and
s2 2 S2 g is an AG-groupoid under the point-wise multiplication of ordered
pairs.
An element (a; b) of an AG-groupoid S1 S 2 is called a regular element
of S1 S 2 if there exist x 2 S1 and m 2 S2 such that (a; b) = ((ax)a; (bm)b)
and S1 S 2 is called regular if all elements of S are regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a weakly regular
element of S1 S 2 if there exist x; y 2 S1 and l; m 2 S2 such that (a; b) =
((ax)(ay); (bl)(bm)) and S1 S 2 is called weakly regular if all elements of
S1 S 2 are weakly regular.
An element (a; b) of an AG-groupoid S1 S 2 is called an intra-regular
element of S1 S 2 if there exist x; y 2 S1 and l; m 2 S2 such that (a; b) =
((xa2 )y; (lb2 )m) and S1 S 2 is called intra-regular if all elements of S1 S 2
are intra-regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a right regular
element of S1 S 2 if there exists x 2 S1 and m 2 S2 such that (a; b) =
(a2 x; b2 m) = ((aa)x; (bb)m) and S1 S 2 is called right regular if all elements
of S1 S 2 are right regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a left regular

(xa2 ; mb2 ) = (x(aa); m(bb)) and S1 S 2 is called left regular if all elements
of S1 S 2 are left regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a left quasi reg-
ular element of S1 S 2 if there exist x; y 2 S1 and l; m 2 S2 such that
(a; b) = ((xa)(ya); (lb)(mb)) and S1 S 2 is called left quasi regular if all
elements of S1 S 2 are left quasi regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a completely
regular element of S if (a; b) is regular, left regular and right regular.
S1 S 2 is called completely regular if it is regular, left and right regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a (2,2)-regular
((a2 x)a2 ; (b2 m)b2 ) and S1 S 2 is called (2; 2)-regular AG-groupoid if all
elements of S1 S 2 are (2; 2)-regular.
An element (a; b) of an AG-groupoid S1 S 2 is called a strongly regular
((ax)a; (bm)b) and ax = xa; bm = mb: S1 S 2 is called strongly regular
AG-groupoid if all elements of S1 S 2 are strongly regular.
Example 81 Let us consider an AG-groupoid S = fa; b; cg in the following
multiplication table.
: a b c
a c c c
b c c a
c c c a
Clearly S is non-commutative and non-associative, because bc 6= cb and

(cc)a 6= c(ca): Note that S has no left identity.
Example 82 Let us consider an AG-groupoid S1 = fa; b; c; d; e; f g with
left identity e and S2 = fg; h; i; j; k; lg with left identity j in the following
Cayley’s tables.
. a b c d e f . g h i j k l
a a a a a a a g g g g g g g
b a b b b b b h g h h h h h
c a b f f d f i g h l l i l
d a b f f c f j g h i j k l
e a b c d e f k g h l l l l
f a b f f f f l g h l l j l
Clearly S1 S 2 is non-commutative and non-associative, because (ed; ik) 6=

(de; ki) and ((de)e; (ik)k) 6= (d(ee); i(kk)):
Lemma 83 If S1 S 2 is a regular, weakly regular, intra-regular, right reg-

ular, left regular, left quasi regular, completely regular, (2; 2)-regular or
strongly regular AG-groupoid; then S1 S 2 = (S1 S 2 )2 :
Proof. Let S1 S 2 be a regular AG-groupoid, then (S1 S 2 )2 S1 S 2 is

obvious. Let (a; b) 2 S1 S 2 where a 2 S1 and b 2 S2 ; then since S1 S 2
is regular so there exists (x; y) 2 S1 S 2 such that (a; b) = ((ax)a; (by)b):
Now by using (2); we have
(a; b) = ((ax)a; (by)b) = ((ax)(by); (ab)) 2 (S1 S 2 )(S1 S 2 ) = (S1 S 2 )2 :
Similarly if S1 S 2 is weakly regular, intra-regular, right regular, left

regular, left quasi regular, completely regular, (2; 2)-regular or strongly
regular, then we can show that S1 S 2 = (S1 S 2 )2 .
The converse is not true in general, because S1 S 2 = (S1 S 2 )2 holds
but S1 S 2 is not regular, weakly regular, intra-regular, right regular, left
regular, left quasi regular, completely regular, (2; 2)-regular and strongly
regular, because (d; k) 2 S1 S 2 is not regular, weakly regular, intra-
regular, right regular, left regular, left quasi regular, completely regular,
(2; 2)-regular and strongly regular.
Theorem 84 If S1 S 2 is an AG-groupoid with left identity (AG -groupoid);

then S1 S 2 is intra-regular if and only if for all (a; b) 2 S1 S 2 ; (a; b) =
((xa)(az); (lb)(bm)) holds for some x; z 2 S1 and l; m 2 S2 :
Proof. Let S1 S 2 be an intra-regular AG-groupoid with left identity (AG -

groupoid), then for any (a; b) 2 S1 S 2 ; there exist x; y 2 S1 and l; k 2 S2
such that (a; b) = ((xa2 )y; (lb2 )k): Now y = uv and k = pq for some
u; v 2 S1 and p; q 2 S2 . Thus we have

(a; b) = ((xa2 )y; (lb2 )k) = ((x(aa))y; (l(bb))k)
= ((a(xa))y; (b(lb))k) = ((y(xa))a; (k(lb))b)
= ((y(xa))((xa2 )y); (k(lb))((lb2 )k))
= (((uv)(xa))((xa2 )y); ((pq)(lb))((lb2 )k))
= (((ax)(vu))((xa2 )y); ((bl)(qp))((lb2 )k))
= (((ax)t)((xa2 )y); ((bl)j)((lb2 )k))
= ((((xa2 )y)t)(ax); (((lb2 )k)j)(bl))
= (((ty)(xa2 ))(ax); ((jk)(lb2 ))(bl))
= (((a2 x)(yt))(ax); ((b2 l)(kj))(bl))
= (((a2 x)s)(ax); ((b2 l)r)(bl))
= (((sx)(aa))(ax); ((rl)(bb))(bl))
= (((aa)(xs))(ax); ((bb)(lr))(bl))
= (((aa)w)(ax); ((bb)n)(bl))
= (((wa)a)(ax); ((nb)b)(bl))
= ((za)(ax); (mb)(bl))
= ((xa)(az); (lb)(bm));
where vu = t; qp = j; yt = s; kj = r; xs = w; lr = n, wa = z and
nb = m for some t; s; w; z 2 S1 and j; r; n; m 2 S2 :
Conversely, let for all (a; b) 2 S1 S 2 ; (a; b) = ((xa)(az); (lb)(bm)) holds
for some x; z 2 S1 and l; m 2 S2 : Now we have
(a; b) = ((xa)(az); (lb)(bm)) = (a((xa)z); b((lb)m))
= (((xa)(az))((xa)z); ((lb)(bm))((lb)m))
= ((a((xa)z))((xa)z); (b((lb)m))((lb)m))
= ((((xa)z)((xa)z))a; (((lb)m)((lb)m))b)
= ((((xa)(xa))(zz))a; (((lb)(lb))(mm))b)
= ((((ax)(ax))(zz))a; (((bl)(bl))(mm))b)
= (((a((ax)x))(zz))a; ((b((bl)l))(mm))b)
= ((((zz)((ax)x))a)a; (((mm)((bl)l))b)b)
= (((z 2 ((ax)x))a)a; ((m2 ((bl)l))b)b)
= ((((ax)(z 2 x))a)a; (((bl)(m2 l))b)b)
= (((((z 2 x)x)a)a)a; ((((m2 l)l)b)b)b)
= ((((x2 z 2 )a)a)a; (((l2 m2 )b)b)b)
= ((a2 (x2 z 2 ))a; (b2 (l2 m2 ))b)
= ((a(x2 z 2 ))(aa); (b(l2 m2 ))(bb))
= ((at)(aa); (bs)(bb));
where x2 z 2 = t and l2 m2 = s for some t 2 S1 ; s 2 S2 :

Now we have
(a; b) = ((at)(aa); (bs)(bb))

= ((((at)(aa))t)(aa); (((bs)(bb))s)(bb))
= ((((aa)(ta))t)(aa); (((bb)(sb))s)(bb))
= (((a2 (ta))t)(aa); ((b2 (sb))s)(bb))
= (((t(ta))a2 )(aa); ((s(sb))b2 )(bb))
= ((ua2 )v; (pb2 )q):
Where t(ta) = u, aa = v; s(sb) = p and bb = q for some u; v 2 S1 ;

p; q 2 S2 : Thus S1 S 2 is intra-regular.

then the following are equivalent.
(i) S1 S 2 is weakly regular.

(ii) S1 S 2 is intra-regular.
Proof. (i) =) (ii) Let S1 S 2 be a weakly regular AG-groupoid with left
identity (AG -groupoid), then for any (a; b) 2 S1 S 2 there exist x; y 2 S1
and l; m 2 S2 such that (a; b) = ((ax)(ay); (bl)(bm)) and x = uv; l = pq for
some u; v 2 S1 ; p; q 2 S2 : Let vu = t 2 S1 and qp = n 2 S2 . Now we have
(a; b) = ((ax)(ay); (bl)(bm))

= ((ya)(xa); (mb)(lb))
= ((ya)((uv)a); (mb)((pq)b))
= ((ya)((av)u); (mb)((bq)p))
= ((av)((ya)u); (bq)((mb)p))
= ((a(ya))(vu); (b(mb))(qp))
= ((a(ya))t; (b(mb))n)
= ((y(aa))t; (m(bb))n) = ((ya2 )t; (mb2 )n):
Thus S1 S 2 is intra-regular.
(ii) =) (i) It is easy.
Theorem 86 If S1 S 2 is an AG-groupoid (AG -groupoid); then the fol-

lowing are equivalent.

(ii) S1 S 2 is right regular.
Proof. (i) =) (ii) Let S1 S 2 be a weakly regular AG-groupoid (AG -
groupoid), then for any (a; b) 2 S1 S 2 there exist x; y 2 S1 and m; n 2 S2
such that (a; b) = ((ax)(ay); (bm)(bn)). Now let xy = t and mn = s for
some t 2 S. Now
(a; b) = ((ax)(ay); (bm)(bn))

= ((aa)(xy); (bb)(mn)) = (a2 t; b2 s):
Thus S1 S 2 is right regular.

(ii) S1 S 2 is left regular.
Proof. (i) =) (ii) Let S1 S 2 be a weakly regular AG-groupoid with left
and m; n 2 S2 such that (a; b) = ((ax)(ay); (bm)(bn)). Now by using (2)
and (3); we have
(a; b) = ((ax)(ay); (bm)(bn)) = ((aa)(xy); (bb)(mn))

= ((yx)(aa); (nm)(bb)) = ((yx)a2 ; (nm)b2 )
= ((ta2 ); (sb2 )) where yx = t; nm = s for some t 2 S1 and s 2 S2 :
Thus S1 S 2 is left regular.

(ii) =) (i) It follows easily:
(ii) S1 S 2 is left quasi regular.
Proof. The proof of this Lemma is straight forward, so is omitted.
Theorem 89 If S1 S 2 is an AG-groupoid with left identity; then the fol-
(i) S1 S 2 is (2; 2)-regular.
(ii) S1 S 2 is completely regular.
Proof. (i) =) (ii) Let S1 S 2 be a (2; 2)-regular AG-groupoid with left
identity, then for (a; b) 2 S1 S 2 there exist x 2 S1 and m 2 S2 such that
(a; b) = ((a2 x)a2 ; (b2 m)b2 ). Now
(a; b) = ((a2 x)a2 ; (b2 m)b2 ) = (ya2 ; nb2 ); where a2 x = y 2 S1 and b2 m = n 2 S2 ;
and by using (3), we have
(a; b) = ((a2 x)(aa); (b2 m)(bb))

= ((aa)(xa2 ); (bb)(mb2 ))
= (a2 z; b2 l); where xa2 = z 2 S1 and mb2 = l 2 S2 :
And we have
(a; b)
= ((a2 x)(aa); (b2 m)(bb))
= ((aa)(xa2 ); (bb)(mb2 ))
= ((aa)((ex)(aa)); (bb)((em)(bb)))
= ((aa)((aa)(xe)); (bb)((bb)(me)))
= ((aa)(a2 t); (bb)(b2 s))
= (((a2 t)a)a; ((b2 s)b)b)
= ((((aa)t)a)a; (((bb)s)b)b)
= ((((ta)a)a)a; (((sb)b)b)b)
= (((aa)(ta))a; ((bb)(sb))b)
= (((at)(aa))a; ((bs)(bb))b)
= ((a((at)a))a; (b((bs)b))b)
= ((ay)a; (bn)b); where xe = t 2 S1 ;
me = s 2 S2 and (at)a = y 2 S1 ; (bs)b = n 2 S2 :
Thus S1 S 2 is left regular, right regular and regular, so S1 S 2 is com-
pletely regular.
(ii) =) (i) Assume that S1 S 2 is a completely regular AG-groupoid
with left identity, then for any (a; b) 2 S1 S 2 there exist x; y; z 2 S1 and
l; m; n 2 S2 such that (a; b) = ((ax)a; (bl)b); (a; b) = (a2 y; b2 m), (a; b) =
(za2 ; nb2 ): Now
(a; b) = ((ax)a; (bl)b)
= (((a2 y)x)(za2 ); ((b2 m)l)(nb2 ))
= (((xy)a2 )(za2 ); ((lm)b2 )(nb2 ))
= (((za2 )a2 )(xy); ((nb2 )b2 )(lm))
= (((a2 a2 )z)(xy); ((b2 b2 )n)(lm))
= (((xy)z)(a2 a2 ); ((lm)n)(b2 b2 ))
= (a2 (((xy)z)a2 ); b2 (((lm)n)b2 ))
= ((ea2 )(((xy)z)a2 ); (eb2 )(((lm)n)b2 ))
= ((a2 ((xy)z))(a2 e); (b2 ((lm)n))(b2 e))
= ((a2 ((xy)z))((aa)e); (b2 ((lm)n))((bb)e))
= ((a2 ((xy)z))((ea)a); (b2 ((lm)n))((eb)b))
= ((a2 ((xy)z))(aa); (b2 ((lm)n))(bb))
= ((a2 t)a2 ; (b2 s)b2 ); where (xy)z = t 2 S1 and (lm)n = s 2 S2 :
This shows that S1 S 2 is (2; 2)-regular.
Lemma 90 Every weakly regular AG-groupoid S1 S 2 with left identity
(AG -groupoid) is regular.
Proof. Assume that S1 S 2 is a weakly regular AG-groupoid with left

and m; n 2 S2 such that (a; b) = ((ax)(ay); (bm)(bn)): Let xy = t 2 S1 ,
t((yx)a) = u 2 S1 and mn = s 2 S2 ; s((nm)b) = l 2 S2 : Now by using (1),
(2); (3) and (4), we have
(a; b) = ((ax)(ay); (bm)(bn)) = (((ay)x)a; ((bn)m)b)

= (((xy)a)a; ((mn)b)b) = ((ta)a; (sb)b)
= ((t((ax)(ay)))a; (s((bm)(bn)))b)
= ((t((aa)(xy)))a; (s((bb)(mn)))b)
= ((t((yx)(aa)))a; (s((nm)(bb)))b)
= ((t(a((yx)a)))a; (s(b((nm)b)))b)
= ((a(t((yx)a)))a; (b(s((nm)b)))b)
= ((au)a; (bl)b):
Thus S1 S 2 is regular.
The converse of above Lemma is not true in general, as can be seen from
the following example.
Example 91 [51] Let us consider an AG-groupoid S1 = f1; 2; 3; 4g with

left identity 3 and S2 = f5; 6; 7; 8g with left identity 6 in the following
Cayley’s tables.
. 1 2 3 4 . 5 6 7 8
1 2 2 4 4 5 6 6 6 6
2 2 2 2 2 6 5 6 7 8
3 1 2 3 4 7 5 6 5 6
4 1 2 1 2 8 6 6 8 8
Clearly S1 S 2 is regular, because (1; 5) = ((1:3):1; (5:6):5); (2; 6) =

((2:1):2; (6:8):6); (3; 7) = ((3:3):3; (7:6):7) and (4; 8) = ((4:1):4; (8:6):8);
but S1 S 2 is not weakly regular, because (1; 5) 2 S1 S 2 is not a weakly
regular element of S1 S 2 .


(ii) S1 S 2 is completely regular.
Proof. (i) =) (ii) It follows easily
Lemma 93 Every strongly regular AG-groupoid S1 S 2 with left identity

(AG -groupoid) is completely regular.
Proof. Assume that S1 S 2 is a strongly regular AG-groupoid with left
identity (AG -groupoid), then for any (a; b) 2 S1 S 2 there exist x 2 S1 ;
y 2 S2 such that (a; b) = ((ax)a; (by)b), ax = xa and by = yb: Now by
using (1), we have
(a; b) = ((ax)a; (by)b) = ((xa)a; (yb)b)

= ((aa)x; (bb)y) = (a2 x; b2 y):
This shows that S1 S 2 is right regular and so S1 S 2 is completely regular.
Note that a completely regular AG-groupoid S1 S 2 need not to be a

strongly regular AG-groupoid, as can be seen from the following example.
Example 94 Let S = fa; b; c; d; e; f; gg be an AG-groupoid with the follow-
ing multiplication table.
. a b c d e f g
a b d f a c e g
b e g b d f a c
c a c e g b d f
d d f a c e g b
e g b d f a c e
f c e g b d f a
g f a c e g b d
Clearly S1 S 2 is completely regular. Indeed, S1 S 2 is regular, as a =

(a:e):a, b = (b:a):b, c = (c:d):c, d = (d:g):d, e = (e:c):e, f = (f:f ):f , g =
(g:b):g, also S1 S 2 is right regular, as a = (a:a):f; b = (b:b):f; c = (c:c):f;
d = (d:d):f; e = (e:e):f; f = (f:f ):f; g = (g:g):f; and S1 S 2 is left regular,
as a = g:(a:a); b = d:(b:b); c = a:(c:c); d = e:(d:d); e = b:(e:e); f = f:(f:f );
g = c:(g:g); but S1 S 2 is not strongly regular, because ax 6= xa for all
a 2 S1 S 2 :
Theorem 95 In an AG-groupoid S1 S 2 with left identity (AG -groupoid);
the following are equivalent.
(iii) S1 S 2 is right regular.
(iv) S1 S 2 is left regular.
(v) S1 S 2 is left quasi regular.
(vi) S1 S 2 is completely regular.

(vii) For all (a; b) 2 S1 S 2 ; there exist x; y 2 S1 and l; m 2 S1 such that
(a; b) = ((xa)(az); (lb)(bm)):
Proof. (i) =) (ii) It follows from above Theorem.
(ii) =) (iii) It follows from above Theorems.
(iii) =) (iv) It follows from above Theorem.
(iv) =) (v) It follows from above Theorem.
(v) =) (vi) It follows from above Theorems.
(vi) =) (i) It follows from above Theorem.
(ii) () (vii) It follows from above Theorem.
Remark 96 Every intra-regular, right regular, left regular, left quasi reg-
ular and completely regular AG-groupoids S1 S 2 with left identity (AG -
groupoids) are regular.
The converse of above is not true in general. Indeed, from above Example
regular AG-groupoid with left identity is not necessarily intra-regular.
Theorem 97 In an AG-groupoid S1 S 2 with left identity; the following
are equivalent.
(iii) S1 S 2 is right regular.
(iv) S1 S 2 is left regular.
(v) S1 S 2 is left quasi regular.
(vi) S1 S 2 is completely regular.
(vii) For all (a; b) 2 S1 S 2 ; there exist x; y 2 S1 and l; m 2 S1 such that
(a; b) = ((xa)(az); (lb)(bm)):
(viii) S1 S 2 is (2; 2)-regular.
Proof. It is easy.
Remark 98 (2; 2)-regular and strongly regular AG-groupoids S1 S 2 with
left identity are regular.
The converse of above is not true in general, as can be seen from above
Example.
Theorem 99 Direct product of regular, weakly regular, intra-regular, right
regular, left regular, left quasi regular, completely regular, (2; 2)-regular and
strongly regular AG -groupoids S1 S 2 becomes semigroups.
4
Ideals in Abel-Grassmann’s
Groupoids
In this chapter we introduce the concept of left, right, bi, quasi, prime
(quasi-prime) semiprime (quasi-semiprime) ideals in AG-groupoids. We in-
troduce m-system in AG-groupoids. We characterize quasi-prime and quasi-
semiprime ideals and …nd their links with m systems. We characterize
ideals in intra-regular AG-groupoids. Then we characterize intra-regular
AG-groupoids using the properties of these ideals.
4.1 Preliminaries
Let S be an AG-groupoid. By an AG-subgroupoid of S; we means a
non-empty subset A of S such that A2 A.
A non-empty subset A of an AG-groupoid S is called a left (right) ideal
of S if SA A (AS A) and it is called a two-sided ideal if it is both
left and a right ideal of S.
A non-empty subset A of an AG-groupoid S is called a generalized
bi-ideal of S if (AS)A A and an AG-subgroupoid A of S is called a
bi-ideal of S if (AS)A A.
A non-empty subset A of an AG-groupoid S is called a quasi-ideal of
S if SA \ AS A:
Note that every one sided ideal of an AG-groupoid S is a quasi-ideal and
right ideal of S is bi-ideal of S.
A non-empty subset A of an AG-groupoid S is called semiprime if
a2 2 A implies a 2 A:
An AG-subgroupoid A of an AG-groupoid S is called a interior ideal of
S if (SA)S A.
An ideal P of an AG-groupoid S is said to be prime if AB P implies
that either A P or B P; where A and B are ideals of S. A left ideal P
of an AG-groupoid S is said to be a quasi-prime if for left ideals A and
B of S such that AB P; we have either A P or B P:
An ideal P of an AG-groupoid S is called strongly irreducible if A \
B P implies either A P or B P , for all ideals A, B and P of S.
If S is an AG-groupoid with left identity e then the principal left ideal
generated by a …xed element “a” is de…ned as hai = Sa = fsa : s 2 Sg.
Clearly, hai is a left ideal of S contains a. Note that if A is an ideal of
S, then A2 is an ideal of S. Also it is easy to verify that A = hAi and
A2 = hA2 i.
If an AG-groupoid S contains left identity e then S = eS S 2 . Therefore
S = S 2 . Also Sa becomes bi-ideal and quasi-ideal of S. Using paramedial,
medial and left invertive law we get
((Sa)S)Sa (SS)(Sa) = (aS)(SS) = (aS)S = (SS)a = Sa,
It is easy to show that (Sa)(Sa) S(Sa). Hence Sa is a bi-ideal of S. Also
S(Sa) \ (Sa)S S(Sa) Sa.
Therefore Sa is a quasi-ideal of S. Also using medial and paramedial laws
and (1), we get
(Sa)2 = (Sa)(Sa) = (SS)a2 = (aa)(SS) = S((aa)S)
= (SS)((aa)S) = (Sa2 )SS = (Sa2 )S.
Therefore Sa2 = a2 S = (Sa2 )S.
Example 100 Let S = f1; 2; 3; 4; 5; 6g, and the binary operation “ ” be
de…ned on S as follows:
1 2 3 4 5 6
1 x x x x x x
2 x x x x x x
3 x x x x x x
4 x x x x x x
5 x x x x x x
6 x 2 x x x x
Where x 2 f1; 3; 4; 5g. Then (S; :) is an AG-groupoid and f2; xg is an ideal
of S.
A subset M of an AG-groupoid S is called an m-system if for all a; b 2 M ,
there exists a1 2 hai, there exists b1 2 hbi such that a1 b1 2 M [50].
Example 101 Let S = f1; 2; 3; 4; 5; 6; 7; 8g, the binary operation “ ” be
de…ned on S as follows:
1 2 3 4 5 6 7 8
1 1 2 4 4 4 4 4 8
2 8 4 4 4 4 4 4 4
3 4 4 4 4 4 4 4 4
4 4 4 4 4 4 4 4 4
5 4 4 4 4 4 4 4 4
6 4 4 4 4 4 4 4 4
7 4 4 4 4 4 4 4 4
8 2 4 4 4 4 4 4 4
Then (S; ) is an AG-groupoid. The set f1; 2; 4; 8g is an m-system in S,
because if 1; 2 2 M , then 4 2< 1 >, 8 2< 2 > and 4 8 = 4 2 M .
Lemma 102 Product of two right ideals of an AG-groupoid with left iden-
tity is an ideal.
Proof. Let S be an AG-groupoid with left identity, therefore S = S 2 . Now
using medial law, we get
(AB)S = (AB)(SS) = (AS)(BS) AB.
Lemma 103 Product of two left ideals of an AG-groupoid with left identity
is a left ideal.
Proof. Let S be an AG-groupoid with left identity, therefore S = S 2 . Now
using medial law, we get
S(AB) = (SS)(AB) = (SA)(SB) AB.
Lemma 104 Let P be a left ideal of an AG-groupoid S with left identity

e, then the following are equivalent,
(i) P is quasi-prime ideal.
(ii) For all left ideals A and B of S: AB = hABi P ) A P or
B P.
(iii) For all left ideals A and B of S: A * P and B * P ) AB * P .
(iv) For all a; b 2 S: haihbi P ) a 2 P or b 2 P .
Proof. (i) , (ii) , (iii) is trivial.
(i) ) (iv)
Let haihbi P , then by (i) either hai P or hbi P , which implies
that either a 2 P or b 2 P .
(iv) ) (ii)
Let AB P . Let a 2 A and b 2 B, then haihbi P , now by (iv) either
a 2 P or b 2 P , which implies that either A P or B P .
Theorem 105 A left ideal P of an AG-groupoid S with left identity is
quasi-prime if and only if SnP is an m-system.
Proof. Let P is quasi-prime ideal of an AG-groupoid S with left identity
and let a; b 2 SnP which implies that a; b 2 = P . Now by lemma 104(iv), we
have haihbi * P and so haihbi SnP . Now let a1 2 hai and b1 2 hbi which
implies that a1 b1 2 SnP . Hence SnP is an m-system.
Conversely, assume that SnP be an m-system. Let a 2 = P and b 2= P , then
a; b 2 SnP . Now there exists a1 in hai and b1 in hbi such that a1 b1 2 SnP .
This implies that a1 b1 2 = P , which further implies that haihbi * P . Hence
by lemma 104(iv), P is a quasi-prime ideal.
Let P be a left ideal of an AG-groupoid S, P is called quasi-semiprime
if for any left ideal A of S such that A2 P , we have A P .
Lemma 106 Let A be a left ideal of an AG-groupoid S with left identity

e, then the following are equivalent,
(i) A is quasi-semiprime.
(ii) For any left ideals I of S: I 2 = hI 2 i A ) I A.
(iii) For any left ideals I of S: I * A ) I 2 * A.
(iv) For all a 2 S: [hai]2 A ) a 2 A.
Proof. (i) , (ii) , (iii) are trivial.
(i) ) (iv)
Let [hai]2 A, then by (i) hai A, which implies that a 2 A.
(iv) ) (ii)
Let I 2 A, if a 2 I, then [hai]2 A, now by (iv) a 2 A, which implies
that I A.
A subset P of an AG-groupoid S with left identity is called an sp-system
if for all a 2 P , there exists a1 , b1 2 hai such that a1 b1 2 P [50].
Lemma 107 Every right ideal of an AG-groupoid S with left identity e is
an sp-system.
Proof. Let I be a right ideal of an AG-groupoid S with left identity e. Now
let i 2 I and s 2 S. Then by left invertive law, we get si = (es)i = (is)e 2
(IS)S I. Therefore I becomes an ideal of S. Also hii = Si SI I.
Now let i1 , i2 2 hii, which implies that i1 i2 2 I. Hence I is an sp-system.
Note that every right ideal of an AG-groupoid S with left identity be-
comes an ideal of S.
Theorem 108 (a) Each m-system is an sp-system.
(b) A left ideal I of an AG-groupoid S is quasi-semiprime if and only if
SnI is an sp-system.
Proof. (a) Let a 2 M , then there exists a1 , b1 2 hai, such that a1 b1 2 M
implying that M is an sp-system.
(b) A left ideal A of an AG-groupoid S with left identity and let a 2 SnA
which implies that a 2 = A. Now let a1 ; b1 2 hai which by lemma 106(iv),
implies that a1 b1 2 [hai]2 but [hai]2 * A. Therefore a1 b1 2 = A. Hence
a1 b1 2 SnA, which shows that SnA is an sp-system.
Conversely, assume that SnA is an sp-system. Let a 2 = A, then a 2 SnA.
Now there exists a1 and b1 in hai, such that a1 b1 2 SnA which implies that
= A, which further implies that [hai]2 * A. Hence by lemma 106(iv),
a1 b1 2
A is a quasi-semiprime ideal.
4.2 Quasi-ideals of Intra-regular Abel-Grassmann’s

Groupoids
Here we begin with examples of intra-regular AG-groupoids.
Example 109 Let S = f1; 2; 3; 4; 5; 6g, then (S; ) is an AG-groupoid with

left identity 5 as given in the following multiplication table:
1 2 3 4 5 6
1 4 5 6 1 2 3
2 3 4 5 6 1 2
3 2 3 4 5 6 1
4 1 2 3 4 5 6
5 6 1 2 3 4 5
6 5 6 1 2 3 4
Clearly (S; ) is intra-regular because, 1 = 3 12 2; 2 = 1 22 5; 3 =

2 32 5; 4 = 4 42 4; 5 = 3 52 6; 6 = 2 62 2.
Example 110 Let S = fa; b; c; d; eg, and the binary operation " " be de-
…ned on S as follows:
1 2 3 4 5 6
1 1 1 1 1 1 1
2 1 2 1 1 1 1
3 1 1 3 4 5 6
4 1 1 6 3 4 5
5 1 1 5 6 3 4
6 1 1 4 5 6 1
Then clearly (S; ) is an AG-groupoid. Also 1 = (1 12 ) 1; 2 = (2 22 ) 2;

3 = (3 32 ) 3; 4 = (3 42 ) 4 and 5 = (4 52 ) 4, 6 = (3 62 ) 6. Therefore
(S; ) is an intra-regular AG-groupoid. It is easy to see that f1g and f1; 2g
are quasi-ideals of S.
In the rest by S we shall mean AG -groupoid such that S = S 2 .
Theorem 111 For S the following conditions are equivalent.

(ii) R \ L = RL; for every semiprime right ideal R and every left ideal
L:
(iii) A = (AS)A; for every quasi-ideal A:
Proof. (i) ) (iii) : Let A be a quasi ideal of S then, A is an ideal of S,
thus (AS)A A:
Now let a 2 A, and since S is intra-regular so there exist elements x; y
in S such that a = (xa2 )y: Now by using medial law with left identity; left
invertive law; medial law and paramedial law; we have
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a

= (y(x((xa2 )y)))a = (y((xa2 )(xy)))a
= ((xa2 )(y(xy)))a = ((x(aa))(y(xy)))a
= ((a(xa))(y(xy)))a = ((ay)((xa)(xy)))a
= ((xa)((ay)(xy)))a = ((xa)((ax)y 2 ))a
= ((y 2 (ax))(ax))a = (a((y 2 (ax))x))a 2 (AS)A:
Hence A = (AS)A:
(iii) ) (ii) : Clearly RL R \ L holds. Now
S(R \ L) \ (R \ L)S = SR \ SL \ RS \ LS = RS \ SL \ SR \ LS
R \ L \ (SR \ LS) R \ L. And
R \ L = ((R \ L)S)(R \ L) = (RS \ LS)(R \ L)
(R \ LS)(R \ L) RL:
Hence R \ L = RL:
(ii) ) (i) : Assume that R \ L = RL for every right ideal R and every
left ideal L of S. Since a2 2 a2 S, which is a right ideal of S and as by given
assumption a2 S is semiprime which implies that a 2 a2 S. Now clearly Sa
is a left ideal of S and a 2 Sa, Therefore by using left invertive law, medial
law, paramedial law and medial law with left identity, we have
a 2 Sa \ a2 S = (Sa)(a2 S) = (Sa)((aa)S) = (Sa)((Sa)(ea))

(Sa)((Sa)(Sa)) = (Sa)((SS)(aa)) (Sa)((SS)(Sa))
= (Sa)((aS)(SS)) = (Sa)((aS)S) = (aS)((Sa)S)
= (a(Sa))(SS) = (a(Sa))S = (S(aa))S = (Sa2 )S:

(ii) For an ideal I and quasi-ideal Q; I \ Q = IQ and I is semiprime.
(iii) For quasi-ideals Q1 and Q2 ; Q1 \ Q2 = Q1 Q2 and Q1 and Q2 are
semiprime.
Proof. (i) =) (iii) : Let Q1 and Q2 be a quasi-ideal of S. Now Q1 and

Q2 become ideals of S. Therefore Q1 Q2 Q1 \ Q2 . Now let a 2 Q1 \ Q2
which implies that a 2 Q1 and a 2 Q2 . For a 2 S there exists x; y in S
such that a = (xa2 )y. Now using (1) and left invertive law, we get
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a 2 (S (SQ1 ))Q2 (SQ1 ) Q2 Q1 Q2 .

This implies that Q1 \ Q2 Q1 Q2 . Hence Q1 \ Q2 = Q1 Q2 . Next we

will show that Q1 and Q2 are semiprime. For this let a2 2 Q1 . Therefore
a = (xa2 )y 2 (SQ1 )S Q1 . Similarly Q2 is semiprime.
(iii) =) (ii) is obvious.
(ii) =) (i) : Obviously Sa is a quasi-ideal contains a and Sa2 is an ideal
contains a2 . By (ii) Sa2 is semiprime so a 2 Sa2 . Therefore by (ii) we get
a 2 Sa2 \ Sa = (Sa2 )(Sa) (Sa2 )S.
(ii) For quasi-ideals Q1 and Q2 , Q1 \ Q2 = (Q1 Q2 )Q1 .
Proof. (i) =) (ii) : Let Q1 and Q2 be quasi-ideals of S. Now Q1 and Q2 be-
come ideals of S. Therefore (Q1 Q2 )Q1 (Q1 S)Q1 Q1 and (Q1 Q2 )Q1
(SQ2 )S Q2 . This implies that (Q1 Q2 )Q1 Q1 \ Q2 . We can easily see
that Q1 \ Q2 becomes an ideal. Now, we get,
2 2
Q1 \ Q2 = (Q1 \ Q2 ) = (Q1 \ Q2 ) (Q1 \ Q2 )
= ((Q1 \ Q2 ) (Q1 \ Q2 )) (Q1 \ Q2 ) (Q1 Q2 )Q1 :
Thus Q1 \ Q2 (Q1 Q2 )Q1 . Hence Q1 \ Q2 = (Q1 Q2 )Q1 .

(ii) =) (i) : Let Q be a quasi-ideal of S, then by (ii), we get Q = Q\Q =
(QQ)Q Q2 Q QQ = Q2 . This implies that Q Q2 therefore Q2 = Q.
Now since Sa is a quasi-ideal, therefore a 2 Sa = (Sa)2 = Sa2 = Sa2 S.
(ii) For quasi-ideal Q and ideal J, Q \ J JQ; and J is semiprime.
Proof. (i) =) (ii) : Assume that Q is a quasi-ideal and J is an ideal of S.
Let a 2 Q \ J; then a 2 Q and a 2 J. For each a 2 S there exists x, y in S
such that a = (xa2 )y. Then using (1) and left invertive law we get,
a = (xa2 )y = (x (aa)) y = (a (xa)) y = (y(xa))a 2 (S(SJ)Q JQ:
Therefore Q \ J JQ. Next let a2 2 J. Thus a = (xa2 )y 2 (SJ)S J.

Hence J is semiprime.
(ii) =) (i) : Since Sa is a quasi and a2 S is a an ideal of S containing a
and a2 respectively. Thus by (ii) J is semiprime so a 2 a2 S: Therefore by
hypothesis, paramedial and medial laws, we get
a 2 Sa \ a2 S (Sa)(a2 S) = (Sa2 )(aS) (Sa2 )S.
Theorem 115 If A is an interior ideal of S, then A2 is also interior ideal.

Proof. Using medial law we immediately obtained the following
SA2 S = ((SS) (AA))(SS) = ((SA) (SA)) (SS)
= ((SA) S) ((SA) S) AA = A2 .

(ii) For quasi-ideal Q, right ideal R and two sided ideal I; (Q \ R) \ I
(QR)I and R; I are semiprime.
(iii) For quasi-ideal Q, right ideal R and right ideal I; (Q\R)\I (QR)I
and R; I are semiprime.
(iv) For quasi-ideal Q, right ideal R and interior ideal I; (Q \ R) \ I
(QR)I and R; I are semiprime.
Proof. (i) =) (iv) : Let a 2 (Q \ R) \ I. This implies that a 2 Q; a 2 R;
a 2 I. Since S is intra-regular therefore for each a 2 S there exists x; y 2 S
such that a = xa2 y. Now using left invertive law, medial law, paramedial
law and (1) we get,
a = xa2 y = (x (aa)) y = (a (xa)) y = a x( xa2 y )) y
= a xa2 (xy) y = y xa2 (xy) a = (y((x(aa)) (xy)) a
= (y ((a(xa)) (xy))) a = ((a (xa)) (y (xy))) a
= (((y (xy)) (xa))a)a 2 (((S (SS)) (SQ))R)I (QR)I.
Therefore (Q \ R) \ I (QR)I. Next let a2 2 R. Then using left invertive
law, we get
a = (xa2 )y = (x(aa))y = (a (xa)) y = (y (xa)) a 2 RT .
This implies that a 2 R. Similarly we can show that I is semiprime.
(iv) =) (iii) =) (ii) : are obvious.
(ii) =) (i) : We know that Sa is a quasi and Sa2 is right as well as two
sided ideal of S containing a and a2 respectively, and by (ii) Sa2 is semi-
prime so a 2 Sa2 . Then by hypothesis and left invertive law, paramedial
and medial laws, we get
a 2 (Sa \ Sa2 ) \ Sa2 = ((Sa)(Sa2 ))Sa2 = ((Sa2 )(Sa2 ))Sa
((Sa2 )S)S = (SS)(Sa2 ) = (a2 S)(SS) = (Sa2 )S:
(ii) For every bi-ideal B and quasi-ideal Q, B \ Q BQ.
(iii) For every generalized bi-ideal B and quasi-ideal Q, B \ Q BQ.
Proof. (i) =) (iii) : Let B is a bi-ideal and Q is a quasi-ideal of S. Let

a 2 B \ Q which implies that a 2 B and a 2 Q. Since S is intra-regular so
for a 2 S there exists x; y 2 S such that a = (xa2 )y. Now B and Q become
ideals of S. Then using (1) and left invertive law, we get
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a 2 (S(SB)Q BQ:
Hence B \ Q BQ.
(ii) =) (i) : Using (ii) we get
a 2 Sa \ Sa Sa2 = Sa2 S.
(ii) For quasi-ideal Q1 , two sided ideal I and quasi-ideal Q2 ; (Q1 \ I) \
Q2 (Q1 I) Q2 , and I is semiprime.
(iii) For quasi-ideal Q1 , right ideal I and quasi ideal Q2 ; (Q1 \ I) \ Q2
(Q1 I) Q2 ; and I is semiprime.
(iv) For quasi-ideal Q1 , interior ideal I and quasi-ideal Q2 (Q1 \ I) \
Q2 (Q1 I) Q2 ; and I is semiprime.
Proof. (i) =) (v) : Let Q1 and Q2 be quasi-ideals and I be an interior ideal
of S respectively. Let a 2 (Q1 \ I) \ Q2 . This implies that a 2 Q1 ; a 2 I
and a 2 Q2 . For a 2 S there exists x; y 2 S such that a = (xa2 )y. Now
Q1 ; Q2 and I become ideals of S. Therefore by left invertive law, medial
law and paramedial law we get,
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = (y(xa))((y(xa))a)
= [afy(xa)g][(xa)y] 2 [Q1 fS(SI)g][(SQ2 )S] (Q1 I) Q2 .
Hence (Q1 \ I) \ Q2 (Q1 I) Q2 : Next let a2 2 I. Then a = (xa2 )y =
2
I I: This implies that a 2 I. Hence that I is semiprime.
(v) =) (iv) =) (iii) =) (ii) are obvious.
(ii) =) (i) : Since Sa is a quasi and Sa2 is an ideal of S containing a
and a2 respectively. Also by (ii) Sa2 is semiprime so a 2 Sa2 . Thus by
using paramedial and medial laws, we get
a 2 (Sa \ Sa2 ) \ Sa ((Sa)(Sa2 ))Sa = ((a2 S)(aS))Sa
= ((a2 S)(SS))(SS) = ((a2 S)S)S = ((SS)a2 )S = (Sa2 )S:
(ii) Every quasi-ideal is idempotent.
(iii) For quasi-ideals A; B; A \ B = AB \ BA.
Proof. (i) =) (iii) : Let A and B be quasi-ideals of S. Thus
AB \ BA AB SB B and AB \ BA BA SA A:
Hence AB \ BA A \ B. Now let a 2 A \ B. This implies that a 2 A and

a 2 B. Since S is intra-regular AG-groupoid so for a in S there exists
x; y 2 S such that a = xa2 y and y = uv for some u; v in S. Then by (1)
and medial law, we get
a = (xa2 )y = (x(aa))y = (a(xa))(uv) = (au)((xa)v) 2 (AS) ((SB)S) AB.
Similarly we can show that a 2 BA. Thus A \ B AB \ BA. Therefore

A \ B = AB \ BA:
(iii) =) (ii) : Let Q be a quasi-ideal of S. Thus by (iii), Q \ Q =
QQ \ QQ. Hence Q = QQ.
(ii) =) (i) : Since Sa is a quasi-ideal of S contains a and by (ii) it is
idempotent therefore by medial law, we have
a 2 Sa = (Sa)2 = (Sa)(Sa) = (SS)a2 = Sa2 = (Sa2 )S:
(ii) For bi-ideal B, two sided ideal I and quasi-ideal Q,(B \ I) \ Q
(BI)Q and I is semiprime.
(iii) For bi-ideal B, right ideal I and quasi-ideal Q,(B \ I) \ Q (BI)Q
and I is semiprime.
(iv) For generalized bi-ideal B, interior ideal I and quasi-ideal Q,(B \
I) \ Q (BI)Q and I is semiprime.
Proof. (i) =) (iv) : Let B be a generalized bi-ideal, I be an interior ideal
and Q be a quasi-ideal of S respectively. Let a 2 (B \ I) \ Q. This implies
that a 2 B; a 2 I and a 2 Q. Since S is intra-regular so for a 2 S there
exists x; y 2 S such that a = (xa2 )y. Now B; I and Q become ideals of S.
Therefore using left invertive law, medial law, paramedial law and (1) we
get,
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = (y(xa))((y(xa))a)

= [afy(xa)g][(xa)y] 2 [BfS(SI)g][(SQ)S] (BI) Q.
Therefore (B \I)\Q (BI)Q. Next let a2 2 I. Then a = (xa2 )y = I 2 I:

This implies that a 2 I.
(iv) =) (iii) =) (ii) are obvious.
(ii) =) (i) : Clearly Sa is both quasi and bi-ideal containing a and Sa2
is two sided ideal contains a2 respectively. Now by (ii) Sa2 is semiprime so
a 2 Sa2 . Therefore using paramedial, medial laws and left invertive law we
get,
a 2 Sa \ Sa2 \ Sa ( Sa)(Sa2 )(Sa) ( a2 S)(aS )(SS)

( a2 S)(SS )(SS) = ((a2 S)S)S = ((SS)a2 )S = (Sa2 )S.
(ii) For quasi-ideals Q and bi-ideal B, Q \ B QB.
(iii) For quasi-ideal Q and generalized bi-ideal B, Q \ B QB.
Proof. (i) =) (iii) : Let Q and B be quasi and generalized bi-ideal of S.
Let a 2 Q \ B. This implies that a 2 Q and a 2 B. Since S is intra-regular
so for a 2 S there exists x; y 2 S such that a = (xa2 )y. Now, Q and B
becomes ideals of S. Therefore using and left invertive law, we get,
a = xa2 y = (x (aa))y = (a (xa))y = (y (xa)) a 2 (S(SQ)B QB.
Thus a 2 QB. Hence Q \ B QB:

(i) =) (ii) : Clearly Sa is both quasi and bi-ideal of S containing a.
Therefore using (ii); paramedial law, medial law we get
a 2 Sa \ Sa (Sa)(Sa) = (Sa2 ) = (Sa2 )S.
(ii) For every quasi-ideal Q of S; Q = (SQ)2 \ (QS)2 .
Proof. (i) =) (ii) : Let Q be any quasi-ideal of S. Now it becomes an
ideal of S. Now using medial law and paramedial law we get
(SQ)2 \ (QS)2 = (SQ)(SQ) \ (QS)(QS) = QQ \ QQ Q:
Now let a 2 Q and since S is intra-regular so there exists x; y 2 S such

that a = (xa2 )y. Then using left invertive law, medial law and paramedial
law, we get
a = (xa2 )y = (a(xa))y = (y(xa))a = (y(xa))((xa2 )y) = (xa2 )((y(xa))y)

= (y(y(xa)))((aa)x) = (aa)((y(y(xa)))x) = (x(y(y(xa))))(aa)
2 S(QQ) = (SS)(QQ) = (SQ)(SQ) = (SQ)2 :
Thus a 2 (SQ)2 . It is easy to see that (SQ)2 = (QS)2 . Therefore a 2

(SQ)2 \ (QS)2 . Thus Q (SQ)2 \ (QS)2 . Hence (SQ)2 \ (QS)2 = Q.
(ii) ) (i) : Clearly Sa is a quasi-ideal containing a. Thus by (ii) and

paramedial law, medial law and left invertive law we get,
a 2 Sa = (S(Sa))2 = ((SS)(Sa))2 = ((aS)(SS))2 = ((SS)a)2

= (Sa)2 = (Sa2 ) = (Sa2 )S.

(ii) For every quasi-ideal of S, Q = (SQ)2 Q \ (QS)2 Q.
Proof. (i) =) (ii) : Let Q be a quasi-ideal of an intra-regular AG-groupoid

S with left identity. Now it becomes an ideal of S. Then obviously
(SQ)2 Q \ (QS)2 Q Q:

that a = (xa2 )y. Then using left invertive law, paramedial law and medial
law, we have,
a = (xa2 )y = (a(xa))y = (y(xa))a = (y(xa))a = (y(x((xa2 )y)))a

= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((xy)(a2 (xy)))a
= (a2 ((xy) (xy)))a = (a (xy))2 a:
2 2 2
Therefore a 2 ((Q (SS)) Q = (QS) Q. This implies that a 2 (QS) Q.
2 2 2
Hence Q (QS) Q. Now since (QS) = (SQ) , thus Q (SQ)2 Q. There-
fore Q (SQ) Q \ (QS) Q. Hence Q = (SQ) Q \ (QS)2 Q:
2 2 2
(ii) ) (i) : Clearly Sa is a quasi-ideal containing a. Therefore by (ii) we

get,
a 2 Sa = (S(Sa))2 (Sa) (Sa)2 (Sa) = (Sa2 )(Sa) (Sa2 )S.

(ii) For any quasi-ideals Q1 and Q2 of S, Q1 Q2 Q2 Q1 and Q1 , Q2
are semiprime.
Proof. (i) =) (ii) : Let Q1 and Q2 be any quasi-ideals of an intra-regular

AG-groupoid S with left identity. Now Q1 and Q2 become ideals of S. Let
a 2 Q1 Q2 . Then a = uv where u 2 Q1 and v 2 Q2 : Now since S in intra-
regular therefore for u and v in S there exists x1 ; x2 ; y1; y2 2 S such that
a = (( x1 u2 y1 )( x2 v 2 y2 )). Using medial law, paramedial law, medial law
and left invertive law, we have

a = (( x1 u2 y1 )( x2 v 2 y2 )) = x1 u2 x2 v 2 (y2 y1 )
= ((x1 (uu)) (x2 (vv))) (y2 y1 ) = ((u(x1 u)) (v(x2 v))) (y2 y1 )
= (((x2 v)(x1 u)) (vu))(y2 y1 ) = (((x2 x1 )(vu)) (vu))(y2 y1 )
= (((vu) (vu))(x2 x1 ))(y2 y1 ) = ((y2 y1 )(x2 x1 )) (((vu) (vu))
= ((y2 y1 )(x2 x1 )) v 2 u2 = ((y2 y1 )v 2 )((x2 x1 )u2 ):
2 ((SS)Q22 ) (SS) Q22 (SQ2 ) (SQ1 ) Q2 Q1 :
2
Thus a 2 Q2 Q1 . Hence Q1 Q2 Q2 Q1 . Let a 2 Q1 . Then since S is
intra-regular so for a 2 S there exists x; y 2 S such that, a = (xa2 )y. Then
using left invertive law, we get
a = (xa2 )y = (x (aa)) y = (a (xa)) y = (y (xa)) a 2 ((SS)Q1 )Q1 Q1 :
Similarly we can show that Q2 semiprime.
(ii) =) (i) : Let Sa be a quasi-ideal of S containing a then by (ii) and
using medial law we get,
a 2 Sa \ Sa = (Sa)(Sa) = (Sa2 ) = (Sa2 )S:
(ii) For any quasi-ideal A and two sided ideal B of S; A \ B = (AB)A
and B is semiprime.
(iii) For any quasi-ideal A and right ideal B of S; A \ B = (AB)A and
B is semiprime.
(iv) For any quasi-ideal A and interior ideal B of S; A; B; A\B = (AB)A
and B is semiprime.
Proof. (i) ) (iv) : Let A and B be a quasi-ideal and an interior ideal of S
respectively. Now A and B are ideals of S. Then (AB)A (AS)A A and
AB)A (SB)S B. Thus (AB)A A \ B. Next let a 2 A \ B; which
implies that a 2 A and a 2 B: Since S is intra-regular so for a there exists
x; y 2 S; such that a = (xa2 )y. Then using left invertive law, we get,
a = (xa2 )y = (a(xa))y = (y(xa))a = (y(xa))a
= (y(x((xa2 )y)))a = (y((xa2 )(xy)))a
= ((xa2 )(y(xy)))a = ((a(xa))(y(xy)))a
= (((y(xy))(xa))a)a = (aa) ((y(xy))(xa))
(AB) (S (SA)) (AB)A:
Thus A \ B = (AB)A. Next to show that B is semiprime let a2 2 B.
Therefore for each a 2 S there exists x; y 2 S such that a = (xa2 )y 2
BB B. Thus a2 2 B. This implies that a 2 B. Hence B is semiprime.

(ii) =) (i) : Since Sa is quasi-ideal and Sa2 be two sided ideal containing
a and a2 respectively. And by (ii) Sa2 is semiprime so a 2 Sa2 . Therefore
using (ii), left invertive law, medial law,and paramedial law, we get
Sa \ Sa2 = ((Sa)(Sa2 )) (Sa) ((SS) (Sa2 )) (SS) = ( a2 S (SS))S
= ( a2 S S)S = ((SS) a2 )S = (Sa2 )S.
(ii) For every left ideal A and B of S, A \ B = (AB) \ (BA).
(iii) For every quasi ideal A and every left ideal B of S, A \ B = (AB) \
(BA).
(iv) For every quasi ideals A and B of S, A \ B = (AB) \ (BA).
Proof. (i) =) (iv) : Let A and B be any generalized bi-ideal of S; then
A and B are ideals of S. Clearly AB A \ B, now A \ B is an ideal and
A \ B = (A \ B)2 . Now A \ B = (A \ B)2 AB. Thus A \ B = AB and
then A \ B = B \ A = BA. Hence A \ B = (AB) \ (BA).
(ii) ) (i) : Since Sa is a left ideal of an AG-groupoid S with left identity
containing a. Therefore by (ii) and medial law we get
Sa \ Sa = (Sa)(Sa) = Sa2 = (Sa2 )S.
(ii) For any quasi-ideals Q and two sided ideal I of S; Q \ I = (QI)Q
and I is semiprime.
(iii) For any quasi-ideals Q and right ideal I of S; Q \ I = (QI)Q and I
is semiprime.
(iv) For any quasi ideals Q and interior ideal I of S; Q \ I = (QI)Q
and I is semiprime.
Proof. (i) ) (v) : Let Q and I be a quasi-ideal and an interior ideal of
S respectively. Now Q and I are ideals of S. Then (QI)Q (QS)Q Q
and (QI)Q (SI)S I. Thus (QI)Q Q \ I. Next let a 2 Q \ I; which
implies that a 2 Q and a 2 I: Since S is intra-regular so for a there exists
x; y 2 S; such that a = (xa2 )y. Then left invertive law, we get,
a = xa2 y = (x (aa))y = (a(xa))y = a x( xa2 y ) y
= a xa2 (xy) y = y xa2 (xy) a = (y ((x(aa)) (xy))) a
= (y ((a(xa)) (xy))) a = ((a (xa)) (y (xy))) a = (((y (xy)) (xa)) a)a
= (aa) ((y(xy)) (xa)) 2 (QI) (S (SS) (SQ) (QI)Q:
Thus Q \ I = (QI)Q. Next to show that I is semiprime let a2 2 I.

Therefore for each a 2 S there exists x; y 2 S such that a = (xa2 )y 2
(SI) S I. Thus a2 2 I. This implies that a 2 I. Hence I is semiprime.
(ii) =) (i) : Since Sa is a quasi-ideal and Sa2 be a two sided ideal
containing a and a2 respectively. And by (ii) Sa2 is semiprime so a 2 Sa2 .
Therefore using (ii), left invertive law, medial law and paramedial we get,
Sa \ Sa2 = ((Sa)(Sa2 )) (Sa) = ((SS) (Sa2 )) (SS) = ( a2 S (SS))S

= ( a2 S S)S = ((SS) a2 )S = (Sa2 )S:
4.3 Characterizations of Ideals in Intra-regular

AG-groupoids
An element a of an AG-groupoid S is called intra-regular if there exist
x; y 2 S such that a = (xa2 )y and S is called intra-regular, if every
element of S is intra-regular.
Example 128 Let us consider an AG-groupoid S = fa; b; c; d; e; f g with
left identity e in the following Clayey’s table.
. a b c d e f
a a a a a a a
b a b b b b b
c a b f f d f
d a b f f c f
e a b c d e f
f a b f f f f
Example 129 Let us consider the set (R; +) of all real numbers under the
binary operation of addition. If we de…ne a b = b a r; where a; b; r 2 R;
then (R; ) becomes an AG-groupoid as,
(a b) c = c (a b) r=c (b a r) r=c b+a+r r=c b+a
and
(c b) a = a (c b) r=a (b c r) r=a b+c+r r=a b + c:
Since (R; +) is commutative so (a b) c = (c b) a and therefore

(R; ) satis…es a left invertive law. It is easy to observe that (R; ) is non-
commutative and non-associative. The same is hold for set of integers and
rationals. Thus (R; ) is an AG-groupoid which is the generalization of an
AG-groupoid given in 1988 (see [39]). Similarly if we de…ne a b = ba 1 r 1 ;

then (Rnf0g; ) becomes an AG-groupoid and the same holds for the set of
integers and rationals. This AG-groupoid is also the generalization of an
AG-groupoid given in 1988 (see [39]).
An element a of an AG-groupoid S is called an intra-regular if there exist
x; y 2 S such that a = (xa2 )y and S is called intra-regular, if every element
of S is intra-regular.
Example 130 Let S = fa; b; c; d; eg be an AG-groupoid with left identity
b in the following multiplication table.
. a b c d e
a a a a a a
b a b c d e
c a e b c d
d a d e b c
e a c d e b
Clearly S is intra-regular because, a = (aa2 )a; b = (cb2 )e; c = (dc2 )e;
d = (cd2 )c; e = (be2 )e:
An element a of an AG-groupoid S with left identity e is called a left
(right) invertible if there exits x 2 S such that xa = e (ax = e) and a is
called invertible if it is both a left and a right invertible. An AG-groupoid
S is called a left (right) invertible if every element of S is a left (right)
invertible and S is called invertible if it is both a left and a right invertible.
Note that in an AG-groupoid S with left identity, S = S 2 :
Theorem 131 Every AG-groupoid S with left identity is an intra-regular
if S is left (right) invertible.
Proof. Let S be a left invertible AG-groupoid with left identity, then for
0 0
a 2 S there exists a 2 S such that a a = e: Now by using left invertive
law, medial law with left identity and medial law, we have
0
a = ea = e(ea) = (a a)(ea) 2 (Sa)(Sa) = (Sa)((SS)a)
= (Sa)((aS)S) = (aS)((Sa)S) = (a(Sa))(SS)
= (a(Sa))S = (S(aa))S = (Sa2 )S:
Which shows that S is intra-regular. Similarly in the case of right invert-
ible.
Theorem 132 An AG-groupoid S is intra-regular if Sa = S or aS = S
holds for all a 2 S.
Proof. Let S be an AG-groupoid such that Sa = S holds for all a 2 S;
then S = S 2 . Let a 2 S, therefore by using medial law, we have
a 2 S = (SS)S = ((Sa)(Sa))S = ((SS)(aa))S (Sa2 )S:
Which shows that S is intra-regular.

Let a 2 S and assume that aS = S holds for all a 2 S; then by using left
invertive law, we have
a 2 S = SS = (aS)S = (SS)a = Sa:
Thus Sa = S holds for all a 2 S, therefore it follows from above that S is

intra-regular.
The converse is not true in general from Example above.
Corollary 133 If S is an AG-groupoid such that aS = S holds for all a

2 S; then Sa = S holds for all a 2 S:
Theorem 134 If S is intra-regular AG-groupoid with left identity, then

(BS)B = B \ S; where B is a bi-(generalized bi-) ideal of S.
Proof. Let S be an intra-regular AG-groupoid with left identity, then

clearly (BS)B B \ S. Now let b 2 B \ S; which implies that b 2 B and
b 2 S: Since S is intra-regular so there exist x; y 2 S such that b = (xb2 )y:
Now by using medial law with left identity; left invertive law; paramedial
law and medial law; we have
b = (x(bb))y = (b(xb))y = (y(xb))b = (y(x((xb2 )y)))b

= (y((xb2 )(xy)))b = ((xb2 )(y(xy)))b = (((xy)y)(b2 x))b
= ((bb)(((xy)y)x))b = ((bb)((xy)(xy)))b = ((bb)(x2 y 2 ))b
= ((y 2 x2 )(bb))b = (b((y 2 x2 )b))b 2 (BS)B:
This shows that (BS)B = B \ S:

The converse is not true in general. For this, let us consider an AG-
groupoid S with left identity e in Example 128: It is easy to see that
fa; b; f g is a bi-(generalized bi-) ideal of S such that (BS)B = B \ S but
S is not an intra-regular because d 2 S is not an intra-regular.
Corollary 135 If S is intra-regular AG-groupoid with left identity, then

(BS)B = B; where B is a bi-(generalized bi-) ideal of S.
Theorem 136 If S is intra-regular AG-groupoid with left identity, then

(SB)S = S \ B; where B is an interior ideal of S.
Proof. Let S be an intra-regular AG-groupoid with left identity, then

clearly (SB)S S \ B. Now let b 2 S \ B; which implies that b 2 S
and b 2 B: Since S is an intra-regular so there exist x; y 2 S such that
b = (xb2 )y: Now by using paramedial law and left invertive law; we have
b = ((ex)(bb))y = ((bb)(xe))y = (((xe)b)b)y 2 (SB)S:
Which shows that (SB)S = S \ B:

The converse is not true in general. It is easy to see that form Example
128 that fa; b; f g is an interior ideal of an AG-groupoid S with left identity
e such that (SB)S = B \ S but S is not an intra-regular because d 2 S is
not an intra-regular.
Corollary 137 If S is intra-regular AG-groupoid with left identity, then

(SB)S = B; where B is an interior ideal of S.
Let S be an AG-groupoid, then ; 6= A S is called semiprime if a2 2

A implies a 2 A:
Theorem 138 An AG-groupoid S with left identity is intra-regular if L [

R = LR; where L and R are the left and right ideals of S respectively such
that R is semiprime.
Proof. Let S be an AG-groupoid with left identity, then clearly Sa and

a2 S are the left and right ideals of S such that a 2 Sa and a2 2 a2 S;
because by using paramedial law, we have
a2 S = (aa)(SS) = (SS)(aa) = Sa2 :
Therefore by given assumption, a 2 a2 S. Now by using left invertive law,

medial law, paramedial law and medial law with left identity, we have
a 2 Sa [ a2 S = (Sa)(a2 S) = (Sa)((aa)S) = (Sa)((Sa)(ea))

(Sa)((Sa)(Sa)) = (Sa)((SS)(aa)) (Sa)((SS)(Sa))
= (Sa)((aS)(SS)) = (Sa)((aS)S) = (aS)((Sa)S)
= (a(Sa))(SS) = (a(Sa))S = (S(aa))S = (Sa2 )S:

The converse is not true in general. In Example 128, the only left and
right ideal of S is fa; bg, where fa; bg is semiprime such that fa; bg[fa; bg =
fa; bgfa; bg but S is not an intra-regular because d 2 S is not an intra-
regular.
Lemma 139 [38] If S is intra-regular regular AG-groupoid, then S = S 2 .
Theorem 140 For a left invertible AG-groupoid S with left identity, the
following conditions are equivalent.
(ii) R \ L = RL; where R and L are any left and right ideals of S
respectively.
Proof. (i) =) (ii) : Assume that S is intra-regular AG-groupoid with left
identity and let a 2 S; then there exist x; y 2 S such that a = (xa2 )y:
Let R and L be any left and right ideals of S respectively, then obviously
RL R \ L: Now let a 2 R \ L implies that a 2 R and a 2 L: Now by
using medial law with left identity, medial law and left invertive law, we
have
a = (xa2 )y 2 (Sa2 )S = (S(aa))S = (a(Sa))S = (a(Sa))(SS)

= (aS)((Sa)S) = (Sa)((aS)S) = (Sa)((SS)a) = (Sa)(Sa)
(SR)(SL) = ((SS)R)(SL) = ((RS)S)(SL) RL:
This shows that R \ L = RL:

(ii) =) (i) : Let S be a left invertible AG-groupoid with left identity,
0 0
then for a 2 S there exists a 2 S such that a a = e: Since a2 S is a right
ideal and also a left ideal of S such that a2 2 a2 S, therefore by using given
assumption, medial law with left identity and left invertive law, we have
a2 2 a2 S \ a2 S = (a2 S)(a2 S) = a2 ((a2 S)S) = a2 ((SS)a2 )

= (aa)(Sa2 ) = ((Sa2 )a)a:
Thus we get, a2 = ((xa2 )a)a for some x 2 S:

Now by using left invertive law, we have
0 0
(aa)a = (((xa2 )a)a)a
0 0
(a a)a = (a a)(((xa2 )a)
a = (xa2 )a:
This shows that S is intra-regular.

Lemma 141 [38] Every two-sided ideal of an intra-regular AG-groupoid S
with left identity is idempotent.
Theorem 142 In an AG-groupoid S with left identity, the following con-
ditions are equivalent.
(ii) A = (SA)2 ; where A is any left ideal of S.
Proof. (i) =) (ii) : Let A be a left ideal of an intra-regular AG-groupoid
S with left identity; then SA A and (SA)2 = SA A: Now A = AA
SA = (SA)2 ; which implies that A = (SA)2 :
(ii) =) (i) : Let A be a left ideal of S; then A = (SA)2 A2 ; which
implies that A is idempotent and by using Lemma 149, S is intra-regular.
Theorem 143 In an intra-regular AG-groupoid S with left identity, the

(i) A is a bi-(generalized bi-) ideal of S.
(ii) (AS)A = A and A2 = A:
Proof. (i) =) (ii) : Let A be a bi-ideal of an intra-regular AG-groupoid S
with left identity; then (AS)A A. Let a 2 A, then since S is intra-regular
so there exist x; y 2 S such that a = (xa2 )y: Now by using medial law with
left identity; left invertive law; medial law and paramedial law; we have

= (y(x((xa2 )y)))a = (y((xa2 )(xy)))a
= ((xa2 )(y(xy)))a = ((x(aa))(y(xy)))a
= ((a(xa))(y(xy)))a = ((ay)((xa)(xy)))a
= ((xa)((ay)(xy)))a = ((xa)((ax)y 2 ))a
= ((y 2 (ax))(ax))a = (a((y 2 (ax))x))a 2 (AS)A:
Thus (AS)A = A holds. Now by using medial law with left identity; left
invertive law; paramedial law and medial law; we have
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = (y(x((xa2 )y)))a

= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((x(aa))(y(xy)))a
= ((a(xa))(y(xy)))a = (((y(xy))(xa))a)a = (((ax)((xy)y))a)a
= (((ax)(y 2 x))a)a = (((ay 2 )(xx))a)a = (((ay 2 )x2 )a)a
= (((x2 y 2 )a)a)a = (((x2 y 2 )((x(aa))y))a)a
= (((x2 y 2 )((a(xa))y))a)a = (((x2 (a(xa)))(y 2 y))a)a
= (((a(x2 (xa)))y 3 )a)a = (((a((xx)(xa)))y 3 )a)a
= (((a((ax)(xx)))y 3 )a)a = ((((ax)(ax2 ))y 3 )a)a
= ((((aa)(xx2 ))y 3 )a)a = (((y 3 x3 )(aa))a)a
= ((a((y 3 x3 )a))a)a ((AS)A)A AA = A2 :
Hence A = A2 holds.

(i) A is a quasi ideal of S.

(ii) SQ \ QS = Q:
Proof. (i) =) (ii) : Let Q be a quasi ideal of an intra-regular AG-groupoid
S with left identity; then SQ \ QS Q. Let q 2 Q, then since S is intra-
regular so there exist x, y 2 S such that q = (xq 2 )y: Let pq 2 SQ; then
by using medial law with left identity; medial law and paramedial law; we
have
pq = p((xq 2 )y) = (xq 2 )(py) = (x(qq))(py) = (q(xq))(py)

= (qp)((xq)y) = (xq)((qp)y) = (y(qp))(qx)
= q((y(qp))x) 2 QS:
Now let qy 2 QS; then by using left invertive law, medial law with left
identity and paramedial law; we have
qp = ((xq 2 )y)p = (py)(xq 2 ) = (py)(x(qq)) = x((py)(qq))

= x((qq)(yp)) = (qq)(x(yp)) = ((x(yp))q)q 2 SQ:
Hence QS = SQ: As by using medial law with left identity and left
invertive law; we have
q = (xq 2 )y = (x(qq))y = (q(xq))y = (y(xq))q 2 SQ:
Thus q 2 SQ \ QS implies that SQ \ QS = Q.


(i) A is an interior ideal of S.

(ii) (SA)S = A:
Proof. (i) =) (ii) : Let A be an interior ideal of an intra-regular AG-
groupoid S with left identity; then (SA)S A. Let a 2 A, then since S is
intra-regular so there exist x; y 2 S such that a = (xa2 )y: Now by using
medial law with left identity, left invertive law and paramedial law; we have
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = (y(xa))((xa2 )y)

= (((xa2 )y)(xa))y = ((ax)(y(xa2 )))y = (((y(xa2 ))x)a)y 2 (SA)S:
Thus (SA)S = A:

(i) A is a (1; 2)-ideal of S.

(ii) (AS)A2 = A and A2 = A .
Proof. (i) =) (ii) : Let A be a (1; 2)-ideal of an intra-regular AG-groupoid
S with left identity; then (AS)A2 A and A2 A. Let a 2 A, then since
S is intra-regular so there exist x; y 2 S such that a = (xa2 )y: Now by
using medial law with left identity, left invertive law and paramedial law;
we have

= (y(x((xa2 )y)))a = (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a
= (((xy)y)(a2 x))a = ((y 2 x)(a2 x))a = (a2 ((y 2 x)x))a
= (a2 (x2 y 2 ))a = (a(x2 y 2 ))a2 = (a(x2 y 2 ))(aa) 2 (AS)A2 :
Thus (AS)A2 = A: Now by using medial law with left identity; left
invertive law; paramedial law and medial law; we have
= (y(xa))((xa2 )y) = (xa2 )((y(xa))y) = (x(aa))((y(xa))y)
= (a(xa))((y(xa))y) = (((y(xa))y)(xa))a = ((ax)(y(y(xa))))a
= ((((xa2 )y)x)(y(y(xa))))a = (((xy)(xa2 ))(y(y(xa))))a
= (((xy)y)((xa2 )(y(xa))))a = ((y 2 x)((x(aa))(y(xa))))a
= ((y 2 x)((xy)((aa)(xa))))a = ((y 2 x)((aa)((xy)(xa))))a
= ((aa)((y 2 x)((xy)(xa))))a = ((aa)((y 2 x)((xx)(ya))))a
= ((((xx)(ya))(y 2 x))(aa))a = ((((ay)(xx))(y 2 x))(aa))a
= ((((x2 y)a)(y 2 x))(aa))a = (((xy 2 )(a(x2 y)))(aa))a
= ((a((xy 2 )(x2 y)))(aa))a = ((a(x3 y 3 ))(aa))a
2 ((AS)A2 )A AA = A2 :
Hence A2 = A:
Lemma 147 [38]Every non empty subset A of an intra-regular AG-groupoid
S with left identity is a left ideal of S if and only if it is a right ideal of S.
(i) A is a (1; 2)-ideal of S.
(ii) A is a two-sided ideal of S:
Proof. (i) =) (ii) : Assume that S is intra-regular AG-groupoid with left
identity and let A be a (1; 2)-ideal of S then, (AS)A2 A: Let a 2 A, then
since S is intra-regular so there exist x; y 2 S such that a = (xa2 )y: Now
by using medial law with left identity; left invertive law and paramedial
law; we have
sa = s((xa2 )y) = (xa2 )(sy) = (x(aa))(sy) = (a(xa))(sy)
= ((sy)(xa))a = ((sy)(xa))((xa2 )y) = (xa2 )(((sy)(xa))y)
= (y((sy)(xa)))(a2 x) = a2 ((y((sy)(xa)))x)
= (aa)((y((sy)(xa)))x) = (x(y((sy)(xa))))(aa)
= (x(y((ax)(ys))))(aa) = (x((ax)(y(ys))))(aa)
= ((ax)(x(y(ys))))(aa) = ((((xa2 )y)x)(x(y(ys))))(aa)
= (((xy)(xa2 ))(x(y(ys))))(aa) = (((a2 x)(yx))(x(y(ys))))(aa)
= ((((yx)x)a2 )(x(y(ys))))(aa) = (((y(ys))x)(a2 ((yx)x)))(aa)
= (((y(ys))x)(a2 (x2 y)))(aa) = (a2 (((y(ys))x)(x2 y)))(aa)
= ((aa)(((y(ys))x)(x2 y)))(aa) = (((x2 y)((y(ys))x))(aa))(aa)
= (a((x2 y)(((y(ys))x)a)))(aa) 2 (AS)A2 A:
Hence A is a left ideal of S and A is a two-sided ideal of S:

(ii) =) (i) : Let A be a two-sided ideal of S. Let y 2 (AS)A2 ; then
y = (as)b2 for some a; b 2 A and s 2 S: Now by using medial law with left
identity; we have
y = (as)b2 = (as)(bb) = b((as)b) 2 AS A:
Hence (AS)A2 A, therefore A is a (1; 2)-ideal of S.

Lemma 149 [38] Let S be an AG-groupoid, then S is intra-regular if and
only if every left ideal of S is idempotent.
Lemma 150 [38]Every non empty subset A of an intra-regular AG-groupoid
S with left identity is a two-sided ideal of S if and only if it is a quasi ideal
of S.
Theorem 151 A two-sided ideal of an intra-regular AG-groupoid S with
left identity is minimal if and only if it is the intersection of two minimal
two-sided ideals of S.
Proof. Let S be intra-regular AG-groupoid and Q be a minimal two-sided
ideal of S, let a 2 Q. As S(Sa) Sa and S(aS) a(SS) = aS; which
shows that Sa and aS are left ideals of S; so Sa and aS are two-sided ideals
of S.
Now
S(Sa \ aS) \ (Sa \ aS)S = S(Sa) \ S(aS) \ (Sa)S \ (aS)S

(Sa \ aS) \ (Sa)S \ Sa Sa \ aS:
This implies that Sa \ aS is a quasi ideal of S; so, Sa \ aS is a two-sided

ideal of S. Also since a 2 Q, we have
Sa \ aS SQ \ QS Q\Q Q.
Now since Q is minimal, so Sa \ aS = Q; where Sa and aS are minimal

two-sided ideals of S, because let I be an two-sided ideal of S such that
I Sa; then I \ aS Sa \ aS Q; which implies that I \ aS = Q: Thus
Q I: Therefore, we have
Sa SQ SI I; gives Sa = I:
Thus Sa is a minimal two-sided ideal of S. Similarly aS is a minimal two-

sided ideal of S:
Conversely, let Q = I \ J be a two-sided ideal of S, where I and J
are minimal two-sided ideals of S; then, Q is a quasi ideal of S, that is
0 0
SQ \ QS Q: Let Q be a two-sided ideal of S such that Q Q, then
0 0 0 0
SQ \ Q S SQ \ QS Q; also SQ SI I and Q S JS J.
Now
0 0 0 0 0 0
S(SQ ) = (SS) (SQ ) = (Q S) (SS) = (Q S)S = (SS) Q = SQ ;
0
which implies that SQ is a left ideal and hence a two-sided ideal. Similarly
0
Q S is a two-sided ideal of S. Since I and J are minimal two-sided ideals
0 0
of S, therefore SQ = I and Q S = J: But Q = I \ J; which implies that,
0 0 0 0
Q = SQ \ Q S Q : This give us Q = Q and hence Q is minimal.
4.4 Characterizations of Intra-regular

AG-groupoids
Example 152 Let S = fa; b; c; d; eg be an AG-groupoid with left identity
b in the following multiplication table.
. a b c d e
a a a a a a
b a b c d e
c a e b c d
d a d e b c
e a c d e b
Clearly S is intra-regular because, a = (aa2 )a; b = (cb2 )e; c = (dc2 )e;

d = (cd2 )c; e = (be2 )e:
a b c d e
a a a a a a
b a b a a a
c a a e c d
d a a d e c
e a a c d e
Then clearly (S; ) is an AG-groupoid. Also a = (aa2 )a; b = (bb2 )b;

c = (ec2 )c; d = (ed2 )d and e = (ee2 )e. Therefore (S; ) is an intra-regular
AG-groupoid. It is easy to see that fag and fa; bg are ideals of S.
Theorem 154 An AG-groupoid S is intra-regular if Sa = S or aS = S

holds for all a 2 S.
Proof. Let S be an AG-groupoid such that Sa = S holds for all a 2 S;

then S = S 2 . Let a 2 S, therefore by using medial law, we have
S = (SS)S = ((Sa)(Sa))S = ((SS)(aa))S (Sa2 )S:


Let a 2 S and assume that aS = S holds for all a 2 S; then by using left
invertive law, we have
S = SS = (aS)S = (SS)a = Sa:
Thus Sa = S holds for all a 2 S, therefore it follows from above that S is

intra-regular.
Lemma 155 Intersection of two ideals of an AG-groupoid with left identity
is either empty or an ideal.
Proof.
(A \ B)S = AS \ BS A \ B.
Lemma 156 Product of two bi-ideals of an AG-groupoid with left identity

is a bi-ideal.
Lemma 157 If I is an ideal of an intra-regular AG-groupoid S with left
identity, then I = I 2 .
Proof. Clearly I 2 I. Now let i 2 I, then since S is intra-regular therefore
there exists x and y in S such that i = (xi2 )y. Then i = (xi2 )y 2 (SI 2 )S
I 2.
Theorem 158 The intersection of two quasi ideals of an AG-groupoid S
is either empty or a quasi ideal of S:
Proof. Let Q1 and Q2 be quasi-ideals of S: Suppose that Q1 \ Q2 is non-
empty, then
S(Q1 \ Q2 ) \ (Q1 \ Q2 )S (SQ1 \ SQ2 ) \ (Q1 S \ Q2 S)

(SQ1 \ Q1 S) \ (SQ2 \ Q2 S)
Q1 \ Q2 .
Hence Q1 \ Q2 is a quasi-ideal of S:
Theorem 159 [38]For an intra-regular AG-groupoid S with left identity
the following statements are equivalent.
(i) A is a left ideal of S.
(ii) A is a right ideal of S.
(iii) A is an ideal of S.
(iv) A is a bi-ideal of S.
(v) A is a generalized bi-ideal of S.
(vi) A is an interior ideal of S.
(vii) A is a quasi-ideal of S.
(viii) AS = A and SA = A.
Theorem 160 Let S be an AG-groupoid with left identity e then the fol-
lowing conditions are equivalent.
(ii) Every left ideal is idempotent.
Proof. (i) =) (ii)
Let L be a left ideal of an intra-regular AG-groupoid S with left identity.
Obviously L2 L. Now let l 2 L. Since S is intra-regular therefore for l
there exists x and y in S such that l = (xl2 )y. Then using left invertive
law, we get
l = (xl2 )y = (l(xl))y = (y(xl))l 2 (S(SL))L L2 .
Therefore L L2 . Hence L = L2 .
(i) =) (ii)
Since Sa is a left ideal contains a. Therefore using (ii) we get, a 2 Sa =
(Sa)2 = Sa2 = (Sa2 )S.
Theorem 161 For an AG-groupoid S with left identity, the following are
equivalent.
(ii) Every quasi-ideal of S is idempotent.
Proof. (i) =) (ii)
Let Q be a quasi-ideal of S. Let a 2 Q which implies that a2 2 Q then
since S is intra-regular so there exist x; y 2 S such that a = (xa2 )y. Now
by theorem 159, Q is an ideal and Q2 becomes an ideal. Therefore
a = (xa2 )y 2 (SQ2 )S Q2 .
Hence Q = Q2 .
(ii) =) (i)
Clearly Sa is a quasi-ideal. Now by (ii) Sa is idempotent. Therefore
a 2 Sa = (Sa)2 but (Sa)2 = (Sa2 )S. Hence a 2 Sa = (Sa2 )S.
equivalent.
(ii) Q = (SQ)2 \ (QS)2 , for every left ideal Q of S:
(iii) Q = (SQ)2 \ (QS)2 , for every quasi-ideal Q of S:
Proof. (i) =) (vi)
Let Q be a quasi-ideal of an intra-regular AG-groupoid S with left iden-
tity so by theorem 159, Q is an ideal and by theorem 161, Q is idempotent,
then medial law we get
(SQ)2 \ (QS)2 = (SQ)(SQ) \ (QS)(QS) = (SS)(QQ) \ (QQ)(SS)

= (SQ) \ (QS) Q:

that a = (xa2 )y. Then using, left invertive law, paramedial law and medial
law, we have

2 S(QQ) = (SS)(QQ) = (SQ)(SQ) = (SQ)2 :
Thus a 2 (SQ)2 . It is easy to see that (SQ)2 = (QS)2 . Therefore a 2

(SQ)2 \ (QS)2 . Thus Q (SQ)2 \ (QS)2 . Hence (SQ)2 \ (QS)2 = Q.
(ii) ) (i)
Let Q be a left ideal of an AG-groupoid S with left identity then by (ii);
Q = (SQ)2 \ (QS)2 (SQ)2 Q2 : Thus Q = Q2 . Hence by theorem 160,
S is intra-regular.
(ii) A (AS)A; for every quasi-ideal A and A = A2 :
Proof. (i) ) (ii)

Let a 2 A, and since S is intra-regular so there exists elements x; y in S
such that a = (xa2 )y. Now using (1) ; left invertive law and medial law; we
have

= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((x(aa))(y(xy)))a
= ((a(xa))(y(xy)))a = ((ay)((xa)(xy)))a = ((xa)((ay)(xy)))a
= ((xa)((ax)y 2 ))a = ((y 2 (ax))(ax))a = (a((y 2 (ax))x))a 2 (AS)A.
Hence A (AS)A. By theorem 159, A becomes an ideal and let c2 2 A.

Now since S in intra-regular so for c there exists u and v in S such that
(uc2 )v. Then
c = (uc2 )v 2 (SA)S A.
Hence A is semiprime.
(ii) ) (i)
It is same as the converse of theorem 161.
(ii) R \ L = RL; for every right ideal R and every left ideal L and R is
semiprime.
Proof. (i) =) (ii)

Let R, L be right and left ideals of an intra-regular AG-groupoid S
with left identity then by theorem 159, R and L become ideals of S and
so RL R \ L. Now R \ L is an ideal and by R \ L = (R \ L)2 . Thus
R\L = (R\L)2 RL. Therefore R\L = RL. Next let r2 2 R. Then since
S is intra-regular therefore for r there exists x and y such that r = (xr2 )y.
Thus
r = (xr2 )y 2 (SR)S R:
Hence R is semiprime.
(ii) =) (i)
Clearly Sa2 is a right ideal contains a2 . Therefore by (ii) a 2 Sa2 . Since
Sa is left ideal and so we get
a 2 Sa2 \ Sa = (Sa2 )(Sa) (Sa2 )S.
Theorem 165 Let S be an AG-groupoid with left identity e; then the fol-
(ii) B = (BS)B; for every bi-ideal B and B = B 2 .
Proof. (i) =) (ii)
Let B is bi-ideal of S then B is an ideal and B = B 2 . Let b 2 B, now since
S is intra-regular therefore for b there exists x; y in S such that b = (xb2 )y.
Also since S = S 2 , therefore for y in S there exists u; v in S such that
y = uv. Now using medial law and left invertive law, we get
b = (xb2 )y = (xb2 )(uv) = (xu)(b2 v) = b2 ((xu)v)

= ((xu)v)b)b = [(xu)v][(xb2 )y])b = (xb2 )[[(xu)v]y])b
= f[x(bb)][(xu)v]y]gb = f[b(xb)][(xu)v]y]gb
= f[y[(xu)v]][(xb)b]gb = f(xb)[y[(xu)v]]b]gb
= fb[y[(xu)v]](bx)gb = fb[b[y[(xu)v]]x]gb (BS)B:
Therefore B = (BS)B.
(ii) =) (i)
Since Sa is a bi-ideal contains a. Therefore using (ii) we get
a 2 Sa = (Sa)2 = Sa2 = (Sa2 )S:
(ii) Every bi-ideal is idempotent.
Proof. It is the part of theorem 165.
(ii) L \ R = LR; for every right ideal R and every left ideal L and R is
semiprime.
Proof. (i) =) (ii)

Let R is a right and L is a left ideal of an intra-regular AG-groupoid
S with left identity. Then by theorem 159, R and L become ideals of S.
Then clearly LR L \ R. Now let a 2 L \ R which implies that a 2 L
and a 2 R. Then since S is intra-regular so for a there exists x; y in S such
that (xa2 )y. Then using and left invertive law we get
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a 2 LR.
Therefore L \ R LR. Hence L \ R = LR.

Let r2 2 R. Now since S in intra-regular therefore for r there exists u
and v in S such that r = (ur2 )v. Thus
r = (ur2 )v 2 (SR)S R.
Hence R is semiprime.
(ii) =) (i)
Clearly Sa2 is a right ideal contains a2 , therefore by (ii) it is semiprime.
Thus a 2 Sa2 . Also we know that Sa is a left ideal of S. Therefore using
paramedial and medial law we get
a 2 Sa \ Sa2 = (Sa)(Sa2 ) = (a2 S)(aS) = (Sa2 )(aS) (Sa2 )S.
equivalent.
(ii) A \ B = (AB)A, for every bi-ideal A and every quasi-ideal B of S.
(iii) A \ B = (AB)A, for every generalized bi-ideal A and every quasi-
ideal B of S.
Proof. (i) ) (iii)
Let A and B be a generalized bi-ideal and quasi-ideal of an intra-regular
AG-groupoid with left identity. Now by theorem 159, A and B are ideals of
S. Then (AB)A (AS)A A and (AB)A (SB)S B; which implies
that (AB)A A \ B. Next let a 2 A \ B; which implies that a 2 A and
a 2 B: Since S is intra-regular so for a there exist x; y 2 S; such that
a = (xa2 )y; then using (1) and left invertive law, we get

= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a == ((a(xa))(y(xy)))a
= (((y(xy))(xa))a)a (((S(SA)B)A (AB)A:
Thus A \ B = (AB)A:
(ii) =) (i)
Since Sa is both bi and quasi-ideal. Therefore by medial law, we get
a 2 Sa \ Sa = ((Sa)(Sa))(Sa) = ((SS)(aa))(Sa)
= (Sa2 )(Sa) (Sa2 )S.
equivalent.
(ii) (A \ B) \ C = (AB)C, for every left ideal A; every two-sided ideal
B and every left ideal C of S and B is semiprime.
(iii) (A \ B) \ C = (AB)C, for every left ideal A; every right ideal B
and every left ideal C of S and B is semiprime.
(iv) (A \ B) \ C = (AB)C, for every left ideal A; every interior ideal B
and every left ideal C of S and B is semiprime.
Proof. (i) ) (iv)
Let S be a intra-regular AG-groupoid with left identity. Let A, B and C
be left, interior and left ideal of S respectively. Now by theorem 159, A, B
and C become ideals of S. Then
(AB)C (AS)S A; (AB)C (SB)S B and(AB)C (SS)C C.
Thus (AB)C (A \ B) \ C. Now let a 2 (A \ B) \ C; which implies

that a 2 A, a 2 B and a 2 C: Now for a there exists x; y 2 S; such that
a = (xa2 )y; then by using (1) and left invertive law, we get

= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((a(xa))(y(xy)))a
= (((y(xy))(xa))a)a (((S(SA)B)C (AB)C:
Therefore (A \ B) \ C (AB)C. Hence (A \ B) \ C = (AB)C.

Next let b2 2 B. Now for b there exists u and v in S such that b = (ub2 )v.
Thus
b = (ub2 )v 2 (SB)S B.
Hence B is semiprime

(ii) =) (i)
Sa is left ideal and Sa2 (contains a2 ) is an ideal. By (ii), Sa2 is semiprime,
therefore a 2 Sa2 . Now using paramedial, medial and left invertive law, we
get
a 2 Sa \ Sa2 \ Sa = ((Sa)(Sa2 ))(Sa) ((SS)(Sa2 ))S

= (((a2 S)(SS))S = (((a2 S)S)S = ((SS)a2 )S = (Sa2 )S.
equivalent.
(ii) A \ B = (AB) \ (BA), for every bi-ideal A and B of S.
(iii) A \ B = (AB) \ (BA), for every bi-ideal A and every generalized
bi-ideal B of S.
(iv) A \ B = (AB) \ (BA), for every generalized bi-ideals A and B of S.
Proof. (i) =) (iv)

Let A and B be any generalized bi-ideal of an intra-regular AG-groupoid
S with left identity; then by theorem 159, A and B are ideals of S. Clearly
AB A \ B, now A \ B is an ideal and A \ B = (A \ B)2 . Now A \ B =
(A \ B)2 AB. Thus A \ B = AB and then A \ B = B \ A = BA: Hence
A \ B = (AB) \ (BA):
(ii) ) (i)
Let B be a ideal of an AG-groupoid S with left identity. Then by (ii)
B \ B = (BB) \ (BB) = B 2 , so by theorem 166, S is intra-regular.
equivalent.
(ii) B \ G = (BG)B; for every bi-ideal B and every quasi-ideal G.
Proof. (i) =) (ii)

Let a 2 B \ G. Now by theorem 159, B and G become ideals of S. Then
using (1) and left invertive law, we get

= (y(xa2 (xy)))a = (xa2 )(y(xy)))a = (a(xa))(y(xy)))a
= (y(xy)(xa))a)a 2 ((S(Sa))a)a ((S(SB))G)B = ((SB)G)B (BG)B.
Therefore B \ G (BG)B.
Next (BG)B (BS)B B and (BG)B (SG)S G. Therefore
(BG)B B \ G. Hence B \ G = (BG)B.
(ii) =) (i)
Sa is both bi and quasi-ideal of an AG-groupoid S with left identity.
Therefore by medial law we get
a 2 Sa \ Sa = ((Sa)(Sa))(Sa) = ((SS)(aa))(Sa)
= (Sa2 )(Sa) (Sa2 )S.
equivalent.
(ii) B \ I = BI(B \ I BI); for every bi-ideal B and every quasi-ideal
I.
Proof. (i) =) (ii)
Let B and I be bi and quasi ideals of an AG-groupoid S with left identity.
Then by theorem 159, B and I become ideals of S. Now clearly BI B \I.
Next let a 2 B \ I. Now since S is intra-regular so for a there exists x; y in
S such that a = (xa2 )y. Now using left invertive law we get

2 (S(SB))I (SB)I BI.
Therefore B \ I BI. Hence B \ I = BI.

(ii) =) (i)
Sa is both bi and quasi-ideal. Therefore by medial law we get
a 2 Sa \ Sa = (Sa)(Sa) = (SS)a2 = Sa2 = (Sa2 )S:
Theorem 173 For an AG-groupoid S with left identity the following con-
ditions are equivalent.
(ii) Every left ideal of S is idempotent.
(iii) A \ B = AB; for every ideals A, B of S and A; B are semiprime.
(iv) A \ B = AB; for every ideal A, every bi-ideal B of S and A; B are
semiprime.
(v) A \ B = AB; for every bi-ideals A; B of S and A; B are semiprime.
(vi) The set of left ideals forms a semilattice structure.
Proof. (i) () (ii)
It is same as theorem 160.
(i) =) (v)
Let A; B are bi-ideals of an intra-regular AG-groupoid S with left iden-
tity. Then by theorem 159, A and B are ideals of S. Now clearly AB
A \ B. Since A \ B is an ideal and (A \ B)2 = A \ B. Thus A \ B =

(A \ B)2 AB. Therefore (A \ B) = AB. Next let a2 2 A. Now for a
there exists x; y in S such that a = (xa2 )y. Thus a = (xa2 )y 2 (SA)S A.
Hence A is semiprime. Similarly we can show that B is semiprime.
(v) =) (i)
Assume that A is a bi-ideal of an AG-groupoid S with left identity then
by (v) A \ A = AA, that is, A = A2 and by theorem 166, S is intra-regular.
(i) =) (vi)
Let LS denote the set of all left ideas of an intra-regular AG-groupoid
S with left identity and let I and J 2 LS . Now by theorem 159, I and
J become ideals of S. Thus IJ I \ J. Now I \ J is an ideal and so
I \ J = (I \ J)2 . Therefore I \ J IJ. Thus I \ J = IJ which clearly
implies that I \ J = JI. Now clearly all elements (ideals) of LS satisfy
left invertive law. Therefore LS form an AG-groupoid. Also IJ = JI and
I = I 2 , for all I and J in LS . But we know that a commutative AG-groupoid
becomes a commutative semigroup. Hence the set of all left ideals that is
LS form a semilattice structure.
(vi) =) (i)
If I is a left ideal of an AG-groupoid S with left identity, then by (vi),
I = I 2 . The rest is same as (ii) =) (i).
(v) =) (iv) =) (iii) are obvious.
(iii) =) (i)
Since Sa2 is an ideal of an AG-groupoid S with left identity. Then by
(iii) it becomes semiprime and since S itself is an ideal, therefore by (iii)
we get
a 2 Sa2 = Sa2 \ S = (Sa2 )S.

AG -groupoids
It is easy to see that every AG-groupoid with left identity becomes an
AG -groupoid but the converse is not true (see the example below)
Example 174 Let S = f1; 2; 3; 4; 5g, the binary operation \ " be de…ned
on S as follows:
1 2 3 4 5
1 1 2 4 4 5
2 5 4 4 4 4
3 4 4 4 4 4
4 4 4 4 4 4
5 2 4 4 4 4
(S; :) is neither commutative nor associative because 5 = 1:5 6= 5:1 = 2

and 2 = (2:1):1 6= 2:(2:1) = 5. Also by AG-test in [48], it is easy to check
that S is an AG -groupoid.
Here we begin with examples of intra-regular AG-groupoids.
Example 175 Let S = f1; 2; 3; 4; 5; 6g, then by AG-test in [48], (S; ) is
an AG-groupoid with left identity 5 as given in the following multiplication
table:
1 2 3 4 5 6
1 5 6 1 2 3 4
2 4 5 6 1 2 3
3 3 4 5 6 1 2
4 2 3 4 5 6 1
5 1 2 3 4 5 6
6 6 1 2 3 4 5
Clearly (S; ) is intra-regular because, 1 = 4 12 2; 2 = 3 22 4; 3 =
2 32 6; 4 = 1 42 2; 5 = 5 52 5; 6 = 3 62 2.
a b c d e
a a a a a a
b a b a a a
c a a e c d
d a a d e c
e a a c d e
Then clearly (S; ) is an AG-groupoid. Also a = (aa2 )a; b = (bb2 )b;

c = (ec2 )c; d = (ed2 )d and e = (ee2 )e. Therefore (S; ) is an intra-regular
AG-groupoid. It is easy to see that fag and fa; bg are ideals of S.
It is easy to note that if S is intra-regular AG-groupoid then S = S 2 .
Lemma 177 Intersection of two ideals of an AG-groupoid is an ideal.
Lemma 178 Product of two bi-ideals of an AG -groupoid is a bi-ideal.
Lemma 179 Let S be an AG -groupoid such that S = S 2 , then every
right ideal is a left ideal.
Proof. Let R be a right ideal of S, then using left invertive law, we get
SR = (SS)R = (RS)S RS R:
Lemma 180 If I is an ideal of an intra-regular AG -groupoid S, then

I = I 2.

Lemma 181 Let S be an AG -groupoid S such that S = S 2 ; then a subset
I of S is a right ideal of S if and only if it is an interior ideal of S.
Corollary 182 Every interior ideal of S becomes a left ideal of S.
Theorem 183 Let S be an intra-regular AG -groupoid, then the following
statements are equivalent.
(i) A is a left ideal of S.
(ii) A is a right ideal of S.
(iii) A is an ideal of S.
(iv) A is a bi-ideal of S.
(v) A is a generalized bi-ideal of S.
(vi) A is an interior ideal of S.
(vii) A is a quasi-ideal of S.
(viii) AS = A and SA = A.
Proof. (i) ) (viii)
Let A be a left ideal of S. Then clearly SA A. Now let a 2 A and since
S is intra-regular for a there exists x; y in S such that a = (xa2 )y. Using
left invertive law we get
a = (xa2 )y = [fx(aa)g]y = [fa(xa)g]y = [fy(xa)g]a 2 SA.
Thus A SA. Therefore SA = A.

Now let a 2 A and s 2 S; since S is an intra-regular, so there exist x,
y 2 S such that a = xa2 y, therefore by left invertive law, we have
as = xa2 y s = ((x (aa)) y) s 2 ((S (AA)) S) S ((S (SA)) S) S ((SA) S) S

= (SS) (SA) = S (SA) = A.
Thus AS A. Next let a 2 A, then since S = S 2 so for y in S there exists

y1 ; y2 in S such that y = y1 y2 . Then using medial law, paramedial law we
get
a = (xa2 )y = (xa2 )(y1 y2 ) = (y2 y1 )(a2 x) = a2 [(y2 y1 )x] 2 AS.
Therefore AS = S:
(viii) ) (vii) ) (vi) ) (v) are same as in [38].
(v) ) (iv)
Let A be a generalized bi-ideal of S. Let a; b 2 A, and since S is intra-
regular so there exist x, y in S such that a = xa2 y; then we have
ab = xa2 y b = [a2 f(y2 y1 )xg]b = [f(y2 y1 )xga2 ]b

= [a(f(y2 y1 )xga)]b 2 (AS) A A.
Hence A is a bi-ideal of S.
(iv) ) (iii) is same as in [38]
(iii) ) (ii) and (ii) ) (i) are obvious.
Lemma 184 In an intra-regular AG -groupoid S, IJ = I \ J, for all

ideals I and J in S.
Proof. Let I and J be ideals of S, then obviously IJ I \J. Since I \J I

2
and I \ J J, then (I \ J) IJ, also I \ J is an ideal of S; so we have
2
I \ J = (I \ J) IJ. Hence IJ = I \ J.
An AG-groupoid S is called totally ordered under inclusion if P and Q
are any ideals of S such that either P Q or Q P .
An ideal P of an AG-groupoid S is called strongly irreducible if A\B P
implies either A P or B P , for all ideals A, B and P of S.
Lemma 185 Every ideal of an intra-regular S is prime if and only if it is

strongly irreducible.
Proof. It is an easy.
Theorem 186 Every ideal of an intra-regular AG-groupoid S is prime if

and only if S is totally ordered under inclusion.
Proof. Assume that every ideal of S is prime. Let P and Q be any ideals
of S, so, P Q = P \ Q, where P \ Q is ideal of S, so is prime, therefore
P Q P \ Q; which implies that P P \ Q or Q P \ Q; which implies
that P Q or Q P . Hence S is totally ordered under inclusion.
Conversely, assume that S is totally ordered under inclusion. Let I, J
and P be any ideals of S such that IJ P . Now without loss of generality
assume that I J then
I = I 2 = II IJ P.
Therefore either I P or J P , which implies that P is prime.
Theorem 187 Let S be an intra-regular AG -groupoid such that S = S 2 ,

then the set of all ideals IS of S, forms a semilattice structure.
Proof. Let A, B 2 IS , since A and B are ideals of S, therefore using medial

law, we have
(AB)S = (AB) (SS) = (AS) (BS) AB.

Also S(AB) = (SS) (AB) = (SA) (SB) AB.
Thus AB is an ideal of S. Hence Is is closed. Also we have, AB = A \ B =

B \ A = BA, which implies that IS is commutative, so is associative. Now
A2 = A, for all A 2 IS . Hence IS is semilattice.
Theorem 188 Let S be an AG -groupoid such that S = S 2 ; then the

(ii) For every generalized bi-ideal B, B = B 2 .
Proof. Assume that S is an intra-regular AG -groupoid and B is a gen-

eralized bi-ideal of S. Let b 2 B, and since S is intra-regular so there exist
c, d in S such that b = cb2 d, then we have
b = cb2 d = fc(bb)gd = fb(cb)gd = fd(cb)gb

= [dfc( cb2 d)g]b = [df cb2 (cd)g]b = [ cb2 fd(cd)g]b
= [f(cd)dg b2 c ]b = [b2 (f(cd)dgc)]b = [(cf(cd)dg) b2 ]b
= [b((cf(cd)dg) b)]b 2 ((BS) B)B BB.
Thus B B 2 . Let a; b 2 B; then ab = [a(f(y2 y1 )xga)]b 2 (BS)B B,

2
therefore B B. Hence B 2 = B.
Conversely, consider the subset Sa of S, then using paramedial law, me-
dial law and left invertive law, we get
((Sa)S)(Sa) S(Sa) = (SS)(Sa) = (aS)S = (SS)a = Sa.
Therefore Sa is a generalized bi-ideal. Now by assumption Sa is idempotent,

so by using medial law, we have
a 2 (Sa) (Sa) = ((Sa) (Sa)) (Sa) = ((SS) (aa)) (Sa) Sa2 (SS) = Sa2 S.
Corollary 189 Let S be an AG -groupoid such that S = S 2 ; then the

(ii) For every bi-ideal B, B = B 2 .
Theorem 190 For an AG -groupoid S, then S is intra-regular if and only

if every ideal I is semiprime.
Proof. (i) =) (ii)

Let S be an intra-regular AG -groupoid. Now let a 2 S such that a2 2 I.
For a 2 S there exists x; y in S such that a = (xa2 )y. Therefore a =
(xa2 )y 2 (SI)S I. Hence I is semiprime.
(ii) =) (i)
Obviously Sa2 is an ideal contains a2 . And by (ii) it is semiprime so
a 2 Sa2 . Therefore a 2 Sa2 = (Sa2 )S. Hence S is intra-regular.
Corollary 191 For an AG -groupoid S, then S is intra-regular if and

only if every right ideal is semiprime.
Theorem 192 For an AG -groupoid S, the following are equivalent.

(ii) For generalized bi-ideals B1 and B2 , B1 \ B2 = (B1 B2 )B1 .
Proof. (i) =) (ii)

Let B1 and B2 be generalized bi-ideals of an intra-regular AG -groupoid
S. Now B1 and B2 become ideals of S. Therefore (B1 B2 )B1 (B1 S)B1
B1 and (B1 B2 )B1 (SB2 )S B2 . This implies that (B1 B2 )B1 B1 \B2 .
Now B1 \ B2 becomes an ideal and we get,
2 2
B1 \ B2 = (B1 \ B2 ) = (B1 \ B2 ) (B1 \ B2 )
= ((B1 \ B2 ) (B1 \ B2 )) (B1 \ B2 ) (B1 B2 )B1 :
Thus B1 \ B2 (B1 B2 )B1 . Hence B1 \ B2 = (B1 B2 )B1 .

(i) =) (ii)
Let B be a bi-ideal of an AG -groupoid S, then using (ii), we get
B = B \ B = (BB)B B 2 B BB = B 2 . Hence by theorem 188, S is
intra-regular.
Corollary 193 For an AG -groupoid S, the following are equivalent.

(ii) For bi-ideals B1 and B2 , B1 \ B2 = (B1 B2 )B1 .
Theorem 194 If A is an interior ideal of an intra-regular AG -groupoid

S such that S = S 2 , then A2 is also interior ideal.
Proof. Using medial law we obtained,
SA2 S = ((SS) (AA))(SS) = ((SA) (SA)) (SS)

= ((SA) S) ((SA) S) AA = A2 .

(ii) Every two sided ideal is semiprime.
(iii) Every right ideal is semiprime.
(iv) Every interior ideal is semiprime.
(v) Every generalized interior ideal is semiprime.
Proof. (i) =) (v)

Let I be a generalized interior ideal of an intra-regular AG -groupoid S.
Let a2 2 I. Then since S is intra-regular so for a 2 S there exists x; y 2 S
such that, a = (xa2 )y. Then a = (xa2 )y 2 (SI)S I.
(ii) =) (i)
It is same as the converse of theorem 190.

(ii) Every two sided ideal is semiprime.
(iii) Every bi-ideal is semiprime.
(iv) Every generalized bi-ideal is semiprime.
Proof. (i) =) (iv)
Let B be any generalized bi-ideal of an intra-regular AG -groupoid S.
Let a2 2 B; since S is intra-regular so for a 2 S there exists x; y 2 S such
that, a = (xa2 )y. No B becomes an ideal of S. Therefore a = (xa2 )y 2
(SB) S B.
(ii) =) (i)
It is same as (ii) =) (i) of theorem 195.
Theorem 197 For an AG -groupoid S such that S = S 2 , the following
are equivalent.
(ii) Every left ideal is idempotent.
(iii) For every left ideal L of S; L = (SL)2 \ (LS)2 .
Proof. (i) =) (ii)
Let L be any left ideal of an intra-regular AG -groupoid S so using
medial law and paramedial law we get
(SL)2 \ (LS)2 = (SL)(SL) \ (LS)(LS) = (SS)(LL) \ (LL)(SS)

= (SS)(LL) \ (SS)(LL) = (SS)(LL) = (SL) (SL) LL L:
Now let a 2 L and since S is intra-regular so there exists x; y 2 S such

that a = (xa2 )y. Then using left invertive law, medial law and paramedial
law, we get

2 S(LL) = (SS)(LL) = (SL)(SL) = (SL)2 :
Thus a 2 (SL)2 . It is easy to see that (SL)2 = (LS)2 . Therefore a 2

(SL)2 \ (LS)2 .
Thus L (SL)2 \ (LS)2 . Hence (SL)2 \ (LS)2 = L.
(ii) ) (i)
Clearly Sa is a left ideal contains a, therefore by (ii) it is idempotent.
Therefore using medial law, we get
a 2 Sa = (Sa)(Sa) = (Sa2 ) = (Sa2 )S.
Theorem 198 For an AG -groupoid S such that S = S 2 , the following

are equivalent.
(ii) For every bi-ideal of S, B = (SB)2 B \ (BS)2 B.
Proof. (i) =) (ii)
Let B be a bi-ideal of an intra-regular AG -groupoid S so by using
medial law and paramedial law we get,
(SB)2 B \ (BS)2 B = ((SB)(SB))B \ ((BS)(BS))B
= ((BB)(SS))B \ ((BB)(SS))B
= (B 2 S 2 )B \ (B 2 S 2 )B = (B 2 S 2 )B
(BS)B B:
Now let a 2 B and since S is intra-regular so there exists x; y 2 S such
that a = (xa2 )y. Then using left invertive law, paramedial law and medial
law, we have,
= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((xy)(a2 (xy)))a
= (a2 ((xy) (xy)))a = (a (xy))2 a:
2 2 2
Therefore a 2 ((B (SS)) B = (BS) B. This implies that a 2 (BS) B.
2 2 2
Hence B (BS) B. Now since (BS) = (SB) , thus B (SB)2 B. There-
fore B (SB) B \ (BS) B. Hence B = (SB)2 B \ (BS)2 B:
2 2
(ii) ) (i)
Let B be a bi-ideal of an AG-groupoid S, then by (ii), medial law, para
medial law, left invertive law and (1), we get
B = (SB)2 B \ (BS)2 B = (SB)2 B = (S 2 B 2 )B = (B 2 S)B
= (BS)(BB) = B[(BS)B] B 2 .
Thus B B 2 but B 2 B. Therefore B = B 2 and hence by corollary
189, S is intra-regular.
Theorem 199 Let S be an AG -groupoid such that S = S 2 ; then the
following are equivalent
(i) S is intra-regular,
(ii) Every ideal of S is semiprime.
(ii) Every quasi-ideal of S is semiprime.
Proof. Let Q be a quasi-ideal of an intra-regular AG -groupoid S and let
a2 2 Q. Then using paramedial and medial laws we get
a = a2 ((y2 y1 )x) = (x(y2 y1 ))a2 2 QS \ SQ Q.
Therefore a 2 Q. Hence Q is semiprime.
Converse is same as (ii) =) (i) of theorem 195.
5
Some Characterizations of
Strongly Regular AG-groupoids
In this chapter, we introduce a new class of AG-groupoids namely strongly
regular and characterize it using its ideals.
5.1 Regularities in AG-groupoids

An AG-groupoid S is said to be regular if for every a in S there exists some
x in S such that a = (ax)a:
An AG-groupoid S is said to be intra-regular if for every a in S there
exists some x; y in S such that a = (xa2 )y.
An AG-groupoid S is said to be strongly regular if for every a in S there
exists some x in S such that a = (ax)a and ax = xa.
Here we begin with examples of AG-groupoids.
Example 200 Let S = f1; 2; 3g, the binary operation “ ” be de…ned on S
as follows:
1 2 3
1 2 2 2
2 2 2 2
3 1 2 1
Clearly (S; ) is an AG-groupoid without left identity.
Example 201 Let S = f1; 2; 3; 4g, the binary operation “ ” be de…ned on
S as follows:
1 2 3 4
1 1 2 3 4
2 4 3 3 3
3 3 3 3 3
4 2 3 3 3
Clearly (S; ) is an AG-groupoid with left identity 1.
as follows:
1 2 3
1 1 2 3
2 3 1 2
3 2 3 1
Clearly (S; ) is a strongly regular AG-groupoid with left identity 1.

Note that every strongly regular AG-groupoid is regular, but converse is
not true, for converse consider the following example.
as follows:
1 2 3
1 1 1 1
2 1 1 3
3 1 2 1
Clearly (S; ) is regular AG-groupoid, but not strongly regular.
Theorem 204 Every strongly regular AG-groupoid is intra-regular.
Proof. Let S be strongly regular AG-groupoid,then for every a 2 S there
exists some x 2 S such that a = (ax)a and ax = xa, then using left
invertive law we get
a = (ax)a = (ax)[(ax)a] = (ax)[(xa)a] = (ax)(a2 x)
= [(a2 x)a]x = [(ax)a2 ]x = (ua2 )x, where u = ax:
Converse of above theorem is not true, for converse consider the following
example.
Example 205 Let S = f1; 2; 3; 4; 5; 6; 7g, the binary operation “ ” be de-
1 2 3 4 5 6 7
1 1 3 5 7 2 4 6
2 4 6 1 3 5 7 2
3 7 2 4 6 1 3 5
4 3 5 7 2 4 6 1
5 6 1 3 5 7 2 4
6 2 4 6 1 3 5 7
7 5 7 2 4 6 1 3
Clearly (S; ) is intra-regular AG-groupoid, but not strongly regular.
5.2 Some Characterizations of Strongly Regular

AG-groupoids
Theorem 206 For an AG-groupoid S with left identity the following are
equivalent,
(i) S is strongly regular,
(ii) L \ A LA and L is strongly regular AG-subgroupoid, where L is
any left ideal and A is any subset of S.
Proof. (i) =) (ii)

Let S be a strongly regular AG-groupoid with left identity. Let a 2 L\A,
now since S is strongly regular so there exists some x 2 S such that a =
(ax)a and ax = xa. Then
a = (ax)a = (xa)a 2 (SL)A LA.
Thus L \ A LA. Let a 2 L, thus a 2 S and since S is strongly regular

so there exists an x in S such that a = (ax)a and ax = xa. Let y = (xa)x,
then using left invertive law, we get
y = (xa)x = (ax)x = x2 a 2 SL L.
Now using left invertive law and (1), we get
ya = [(xa)x]a = (ax)(xa) = (xa)(ax) = a[(xa)x] = ay.
Now using left invertive law we get

= a2 [(ax)x] = (aa)[(xa)x] = (aa)y = (ya)a = (ay)a:
Therefore L is strongly regular.

(ii) =) (i)
Since S itself is a left ideal, therefore by assumption S is strongly regular.
equivalent,
(ii) B \ A BA and B is strongly regular AG-subgroupoid, where B is
any bi ideal and A is any subset of S.
Proof. (i) =) (ii)

Let S be a strongly regular AG-groupoid with left identity. Let a 2 B \A,
(ax)a and ax = xa. Then using left invertive law, we get
a = (ax)a = [f(ax)agx]a = [(xa)(ax)]a

= [(ax)(xa)]a = [f(xa)xga]a = [f(xf(ax)ag)xga]a
= [ff(ax)(xa)gxga]a = [ffx(xa)g(ax)ga]a = [(affx(xa)gxg)a]a
= [(at)a]a 2 [(BS)B]A BA; where t = x(xa):
Thus B \ A BA. Let a 2 B, thus a 2 S and since S is strongly regular

so there exists an x in S such that a = (ax)a and ax = xa. Let y = (xa)x,
then using left invertive law, paramedial and medial law, we get
y = (xa)x = (ax)x = x2 a = x2 [(ax)a] = x2 [(xa)a] = x2 (a2 x)

= a2 (x2 x) = a2 t = (aa)t = (ta)a = [tf(ax)ag]a = [tf(xa)ag]a
= [t(a2 x)]a = [a2 (tx)]a = [(aa)(tx)]a = [(xt)(aa)]a = [af(xt)ag]a
= (av)a 2 (BS)B B; where t = (x2 x) and v = (xt)a:
Now using left invertive law, we get
Now using left invertive law we get

= a2 [(ax)x] = (aa)[(xa)x] = (aa)y = (ya)a = (ay)a.
Therefore B is strongly regular.

(ii) =) (i)
Since S itself is a bi ideal, therefore by assumption S is strongly regular.
equivalent,
(ii) Q \ A QA and Q is strongly regular AG-subgroupoid, where Q is
any quasi ideal and A is any subset of S:
Proof. (i) =) (ii)

Let S be a strongly regular AG-groupoid with left identity. Let a 2 Q\A,
(ax)a and ax = xa. Now using left invertive law, we get
ax = [(ax)a]x = [(xa)a]x = (a2 x)x = x2 a2 = x2 (aa)

= a(x2 a) 2 QS.
ax = [(ax)a]x = (xa)(ax) = (ax)(xa) = [(xa)x]a 2 SQ:
Thus ax 2 QS \ SQ Q.
Also a = (ax)a 2 QA. Let a 2 Q, thus a 2 S and since S is strongly
regular so there exists an x in S such that a = (ax)a and ax = xa. Let
y = (xa)x, then using left invertive law, paramedial, medial law, we get
y = (xa)x = (ax)x = x2 a 2 SQ;
and
y = (xa)x = (xa)(ex) = (xe)(ax) = a[(xe)x] 2 QS:
Thus y 2 QS \ SQ Q. Now using left invertive law and (1), we get

= a2 [(ax)x] = (aa)[(xa)x] = (aa)y = (ya)a = (ay)a:
Therefore Q is strongly regular.
(ii) =) (i)
Since S itself is a quasi ideal, therefore by assumption S is strongly
regular.
Theorem 209 Let S be a strongly regular AG-groupoid with left identity.
Then, for every a 2 S; there exists y 2 S such that a = (ay)a, y = (ya)y
and ay = ya:
Proof. Let a 2 S; since S is strongly regular, there exists x 2 S such that
a = (ax)a and ax = xa. Now using paramedial law and medial law,
we get
a = (ax)a = (xa)a = [xf(ax)ag]a = [xf(ax)(ea)g]a
= [xf(ae)(xa)g]a = [(ae)fx(xa)g]a = [(ae)f(ex)(ax)g]a
= [(ae)f(xa)(xe)g]a = [(xa)f(ae)(xe)g]a = [(xa)f(ex)(ea)g]a
= [(xa)(xa)]a = [(ax)(ax)]a = [af(ax)xg]a = [af(xa)xg]a
= (ay)a, where y = (xa)x:
Now using and left invertive law, we get
y = (xa)x = [xf(ax)ag]x = [(ax)(xa)]x
= [f(xa)xga]x = (ya)x = [yf(ax)ag]x
= [xf(ax)ag]y = [xf(ax)ag]y = [(ax)(xa)]y
= [f(xa)xga]y = (ya)y:
ay = a[(xa)x] = (xa)(ax) = (ax)(xa) = [(xa)x]a = ya:
equivalent,
(ii) S is left regular, right regular and (Sa)S is a strongly regular AG-
subgroupoid, of S for every a 2 S:
(iii) For every a 2 S; we have a 2 aS and (Sa)S is a strongly regular
AG-subgroupoid, of S:
Proof. (i) =) (ii)

Let a 2 S, and S is strongly regular so there exists some x 2 S such that
a = (ax)a and ax = xa. Now left invertive law ,we get
a = (ax)a = (xa)a = a2 x:
This implies that S is right regular. Now using medial law and paramedial
law, we get
a = (ax)a = (ax)[(ax)a] = [a(ax)](xa)

= [a(xa)](xa) = [x(aa)](xa) = (xa2 )(xa)
= [x(aa)](xa) = [(ex)(aa)](xa) = [(aa)(xe)](xa)
= [a2 (xe)](xa) = [(xa)(xe)]a2 = ua2 ; where u = [(xa)(xe)]:
Let b 2 (Sa)S S, thus b 2 S; and since S is strongly regular, so there

exist x1 2 S; such that b = (bx1 )b and x1 = (x1 b)x1 and bx1 = x1 b; since
b 2 (Sa)S ) b = (za)t; for some z; t 2 S. Using paramedial, medial law,
left invertive law, we get
x1 = (x1 b)x1 = (x1 b)(ex1 ) = (x1 e)(bx1 ) = b[(x1 e)x1 ] = bu

= [(za)t]u = (ut)(za) = (az)(tu) = [(tu)z]a = va = v(a2 x)
= a2 (vx) = (aa)(vx) 2 (Sa)S; where u = (x1 e)x1 and v = (tu)z:
This shows that (Sa)S is strongly regular.

(ii) =) (iii)
Let a 2 S, and S is left regular so there exists some y 2 S such that
a = ya2 .
Now using (1), we get
a = ya2 = y(aa) = a(ya) 2 a(SS) = aS:
(iii) =) (i)
Let a 2 aS so there exists some t 2 S such that a = at, also a 2 Sa so
there exists some z 2 S such that a = za.
Now
a = at = (za)t 2 (Sa)S;
and as (Sa)S strongly regular so there exists some x in S such that
a = (ax)a and ax = xa. So S is strongly regular.
equivalent,
(ii) (Sa)S is strongly regular and S is left duo.
Proof. (i) =) (ii)

Let a 2 (Sa)S, so a 2 S and since S is strongly regular so there exists
some x 2 S such that a = (ax)a and ax = xa. Let y = (xa)x for any y 2 S.
Now using (1) and left invertive law ,we get
y = (xa)x = [xf(ax)ag]x = [(ax)(xa)]x
= [f(xa)xga]x = (ya)x 2 (Sa)S.
Now using paramedial law,medial law, we get
= (ay)a;
and using (1) and left invertive law, we get
ay = a[(xa)x] = (xa)(ax) = (ax)(xa) = [(xa)x]a = ya:
This shows that (Sa)S is strongly regular.
Let L be any left ideal in S ) SL L. Let a 2 L; s 2 S. Since S is
strongly regular, so there exists some x 2 S, such that, a = (ax)a and
ax = xa. Now as 2 LS
as = [(ax)a]s = [(xa)a]s = (a2 x)s = (sx)a2 = (sx)(aa) 2 S(SL) SL L:
This shows that L is also right ideal and S is left duo.
(ii) =) (i)
Using medial and paramedial laws we get (Sa)(SS) = (SS)(aS) =
(Sa)S. Now since S is left duo, so aS Sa. Also we can show that Sa aS.
Thus Sa = aS. Now let a 2 S; also a 2 Sa = aS ) a = ta and a = av for
some t; v 2 S: Now
a = av = (ta)v 2 (Sa)S:
As (Sa)S is strongly regular, so there exists some u 2 (Sa)S; such that
a = (au)a and au = ua. Hence S is regular.
equivalent,
(ii) Sa is strongly regular for all a in S.
Proof. (i) =) (ii)
Let a 2 Sa, so a 2 S and S is strongly regular so there exists some x 2 S
such that a = (ax)a and ax = xa: Let y = (xa)x for some y 2 S: Now
using left invertive law we get
y = (xa)x = (ax)x = x2 a 2 Sa.
Now using paramedial law,medial law, we get

= (ay)a;
and using left invertive law, we get
ay = a[(xa)x] = (xa)(ax) = (ax)(xa)

= [(xa)x]a = ya:
Which implies that Sa is strongly regular.

(ii) =) (i)
Let a 2 S; so a 2 Sa and Sa is strongly regular which implies S is
strongly regular.
6
Fuzzy Ideals in
Abel-Grassmann’s Groupoids
In this chapter we introduce the fuzzy ideals in AG-groupoids and discuss
their related properties.
A fuzzy subset f of an AG-groupoid S is called a fuzzy AG-subgroupoid
of S if f (xy) f (x) ^ f (y) for all x, y 2 S: A fuzzy subset f of an
AG-groupoid S is called a fuzzy left (right) ideal of S if f (xy) f (y)
(f (xy) f (x)) for all x, y 2 S. A fuzzy subset f of an AG-groupoid S
is called a fuzzy two-sided ideal of S if it is both a fuzzy left and a fuzzy
right ideal of S. A fuzzy subset f of an AG-groupoid S is called a fuzzy
quasi-ideal of S if f S \ S f f . A fuzzy subset f of an AG-groupoid S
is called a fuzzy generalized bi-ideal of S if f ((xa)y) f (x) ^ f (y), for all
x, a and y 2 S. A fuzzy AG-subgroupoid f of an AG-groupoid S is called a
fuzzy bi-ideal of S if f ((xa)y) f (x) ^ f (y), for all x, a and y 2 S. A fuzzy
AG-subgroupoid f of an AG-groupoid S is called a fuzzy interior ideal of
S if f ((xa)y) f (a), for all x, a and y 2 S.
Let f and g be any two fuzzy subsets of an AG-groupoid S, then the
product f g is de…ned by,
8 _
< ff (b) ^ g(c)g , if there exist b; c 2 S, such that a = bc:
(f g) (a) = a=bc
:
0; otherwise.
The symbols f \ g and f [ g will means the following fuzzy subsets of S
(f \ g)(x) = minff (x); g(x)g = f (x) ^ g(x); for all x in S
and
(f [ g)(x) = maxff (x); g(x)g = f (x) _ g(x); for all x in S:
The proof of the following three lemma’s are same as in [37].
Lemma 213 Let f be a fuzzy subset of an AG-groupoid S. Then the

following properties hold.
(i) f is a fuzzy AG-subgroupoid of S if and only if f f f .

(ii) f is a fuzzy left(right) ideal of S if and only if S f f (f S f ).
(iii) f is a fuzzy two-sided ideal of S if and only if S f f and f S f.
Lemma 214 Let f be a fuzzy AG-subgroupoid of an AG-groupoid S. Then

f is a fuzzy bi-ideal of S if and only if (f S) f f .
Lemma 215 Let f be a fuzzy AG-subgroupoid of an AG-groupoid S. Then

f is a fuzzy interior ideal of S if and only if (S f ) S f .
The principal left, right and two-sided ideals of an AG-groupoid S is

denoted by L[a2 ], R[a2 ] and J[a2 ]. Note that the principal left, right and
two-sided ideals generated by a2 are equals, that is,
L[a2 ] = R[a2 ] = J[a2 ] = a2 S = Sa2 S = Sa2 = fsa2 : s 2 Sg:

The characteristic function CA for a subset A of an AG-groupoid S is
de…ned by
1, if x 2 A;
CA (x) =
0, if x 2
= A:
The proof of the following three lemma’s are same as in [29].
Lemma 216 Let A be a non-empty subset of an AG-groupoid S. Then the

following properties hold.
(i) A is an AG-subgroupoid if and only if CA is a fuzzy AG-subgroupoid

of S.
(ii) A is a left(right, two-sided) ideal of S if and only if CA is a fuzzy
left(right, two-sided) of S.
(iii) A is a bi-ideal of S if and only if CA is a fuzzy bi-ideal of S.
Lemma 217 Let A be a non-empty subset of an AG-groupoid S. Then A

is a bi-ideal of S if and only if CA is a fuzzy bi-ideal of S.
Lemma 218 Let A be a non-empty subset of an AG-groupoid S. Then A

is an interior ideal of S if and only if CA is a fuzzy interior ideal of S.
Example 219 Let S = f1; 2; 3; 4g, the binary operation " " be de…ned on
S as follows:
1 2 3 4
1 4 4 2 2
2 4 4 4 4
3 4 4 2 4
4 4 4 4 4
Then (S; ) is an AG-groupoid.
Let us denote the set of all fuzzy subsets of an AG-groupoid S by F (S).

Note that S(x) = 1, for all x 2 S.
Lemma 220 If S is an AG-groupoid with left identity e, then F (S) is an

AG-groupoid with left identity S.
Proof. Let f be any subset of F (S), and for any a 2 S, since e is left
identity of S. So, a = ea, then we have
_ _
(S f )(a) = fS(e) ^ f (a)g = f1 ^ f (a)g = f (a):
a=ea a=ea
Now for uniqueness, suppose S and S 0 be the two left identities of F (S),
then S S 0 = S 0 , and S 0 S = S. Now by using (1), we have S = S 0 S =
(S 0 S 0 ) S = (S S 0 ) S 0 = S 0 S 0 = S 0 .
Lemma 221 In an AG-groupoid F (S), every right identity S is a unique

left identity.
Proof. Let f be any subset of F (S), since S is a right identity of F (S),

then f S = f . Now we have
S f = (S S) f = (f S) S = f S = f.
Lemma 222 An AG-groupoid F (S) with right identity is a commutative

semigroup.
Proof. Since F (S) is an AG-groupoid with right identity S. So by lemma

221, S is left identity of F (S). Let f , g and h 2 F (S), then, we have
f g = (S f ) g = (g f ) S = g f , and
(f g) h = (h g) f = (g h) f = f (g h).
6.1 Inverses in AG-groupoids

Let f be any fuzzy subset of an AG-groupoid S with left identity. A fuzzy
subset f 0 of S is called left(right) inverse of f , if f 0 f = S(f f 0 = S). f 0
is said to be inverse of f if it is both left inverse and right inverse.
Lemma 223 Every right inverse in an AG-groupoid F (S), is left inverse.
Proof. Let f 0 and f be any fuzzy subsets of S and f 0 is the right inverse
of f . Then by using (1), we have
f 0 f = (S f 0 ) f = (f f 0) S = S S = S,
which implies that f 0 is left inverse of f . Now for uniqueness, let f 0 and
f be the two left inverses of f . So f 0 f = S and f 00 f = S. Now by
00
using (1), we have

f 00 = S f 00 = (f 0 f ) f 00 = (f 00 f ) f 0 = S f 0 = f 0.
Let f , g and h be any fuzzy subsets of an AG-groupoid S, then F (S) is

called left (right) cancellative AG-groupoid if f g = f h (g f = h f )
implies that g = h, and F (S) is called a cancellative AG-groupoid if it is
both right and left cancellative.
Lemma 224 A left cancellative AG-groupoid F (S) is a cancellative AG-
groupoid.
Proof. Let F (S) be left cancellative and f , g and h be any fuzzy subsets
of an AG-groupoid S. Now let g f = h f , which implies that (g f ) k =
(h f ) k, where k 2 F (S), now we have (k f ) g = (k f ) h, which
implies that g = h.
Lemma 225 A right cancellative AG-groupoid F (S) with left identity S
is a cancellative AG-groupoid.
Proof. Let f , g , h and k be any fuzzy subsets of an AG-groupoid S.
Let F (S) is right cancellative then g f = h f implies that g = f . Let
f g = f h which implies that g f = h f in [29] which implies that
g = f.
Lemma 226 An AG-groupoid F (S) is a semigroup if and only if f (g
h) = (h g) f . where f , g, h and k are fuzzy subsets of S.
Proof. Let f (g h) = (h g) f , holds for all fuzzy subsets f , g, h and
k of S. Then by using (1), we have f (g h) = (h g) f = (f g) h.
Conversely, suppose that F (S) is a semigroup, then it is easy to see that
f (g h) = (f g) h.
Lemma 227 If f and g be any fuzzy bi-ideals of an AG-groupoid S with
left identity, then f g and g f are fuzzy bi-ideals of S.
Proof. Let f and g be any fuzzy bi-ideals of an AG-groupoid S with left
identity e, then
(f g) (f g) = (f f ) (g g) f g, and
((f g) S) (f g) = ((f g) (S S)) (f g)
= ((f S) (g S)) (f g)
= ((f S) f) ((g S) g) f g.
Hence f g is a fuzzy bi-ideal of S. Similarly g f is a fuzzy bi-ideal of S.
Lemma 228 Every fuzzy ideal of an AG-groupoid S, is a fuzzy bi-ideal

and a fuzzy interior ideal of S.
Proof. Let S be an AG-groupoid and f be any fuzzy ideal of S, then for a,
b 2 S, we have f (ab) f (a) and f (ab) f (b), therefore f (ab) f (a)^f (b),
which implies that f is a fuzzy AG-subgroupoid. Now for any x, y, z 2 S,
we have f ((xy)z) f (xy) f (x), and f ((xy)z) f (z), which implies that
f ((xy)z) f (x) ^ f (z). Hence f is a fuzzy bi-ideal. Similarly it is easy to
see that f ((xa)y) f (a).
Lemma 229 Let f be a fuzzy subset of a completely regular AG-groupoid
S with left identity, then the following are equivalent.
(i) f is a fuzzy ideal of S.
(ii) f is a fuzzy interior ideal of S.
Proof. (i) ) (ii), it is obvious.
(ii) ) (i)
Since S is a completely regular AG-groupoid so for all a, b 2 S there
exist x, y 2 S such that a = (ax)a and ax = xa, b = (by)b and by = yb,
now by using we have
f (ab) = f (((ax)a)b) = f ((ba)(ax)) f (a).
Now we get
f (ab) = f (a((by)b)) = f (a((yb)b)) = f ((yb)(ab)) f (b).
Theorem 230 Every fuzzy generalized bi-ideal of a completely regular AG-

groupoid S with left identity, is a fuzzy bi-ideal of S.
Proof. Let f be any fuzzy generalized bi-ideal of an AG-groupoid S. Then,
since S is completely regular, so for each a 2 S there exist x 2 S such that
a = (ax)a and ax = xa. Thus
f (ab) = f (((ax)a)b) = f (((ax)(ea))b) = f (((xa)(ea))b)
= f (((xe)(aa))b) = f ((a((xe)a))b) f (a) ^ f (b):
Theorem 231 Let f , g and h be any fuzzy subset of an AG-groupoid S,

(i) f (g [ h) = (f g) [ (f h); (g [ h) f = (g f ) [ (h f ).
(ii) f (g \ h) = (f g) \ (f h); (g \ h) f = (g f ) \ (h f ).
Lemma 232 Let f , g and h be any fuzzy subsets of an AG-groupoid S, if
f g, then f h g h and h f h g.

A subset P of an AG-groupoid S is called semiprime, if for all a 2 S,
a2 2 P implies a 2 P .
A fuzzy subset f of an AG-groupoid S is called fuzzy semiprime, if f (a)
f (a2 ), for all a 2 S.
6.2 Fuzzy Semiprime Ideals

Lemma 233 In an intra-regular AG-groupoid S, every fuzzy interior ideal
is fuzzy semiprime.
Proof. Since S intra-regular so for a 2 S there exist x, y 2 S such that
a = (xa2 )y, so we have
f (a) = f ((xa2 )y) f (a2 ).
Theorem 234 A non-empty subset A of an AG-groupoid S, is semiprime

if and only if the characteristic function CA of A is fuzzy semiprime.
Proof. Let a2 2 A, since A is semiprime so a 2 A, hence CA (a) = 1 =
CA (a2 ). Also if a2 2
= A, then CA (a) 0 = CA (a2 ). In both cases, we have
2
CA (a) CA (a ) for all a 2 S, which implies that CA is fuzzy semiprime.
Conversely, assume that a2 2 A, since CA is a fuzzy semiprime, so we
have CA (a) CA (a2 ) = 1, and so CA (a) = 1, which implies that a 2 A.
Theorem 235 For any fuzzy AG-subgroupoid f of an AG-groupoid S, the
following are equivalent.
(i) f is a fuzzy semiprime.
(ii) f (a) = f (a2 ), for all a 2 S.
Proof. (i) ) (ii)
Let a 2 S, then since f is a fuzzy AG-subgroupoid of S, so we have
f (a) f (a2 ) = f (aa) f (a) ^ f (a) = f (a).
(ii) ) (i), it is obvious.

An element a of an AG-groupoid S is called intra-regular if there exists
elements x, y 2 S such that a = (xa2 )y. An AG-groupoid S is called
intra-regular if every element of S is intra-regular.
equivalent.
(ii) f (a) = f (a2 ), for all fuzzy two sided ideal f of S, for all a 2 S.
(iii) f (a) = f (a2 ), for all fuzzy interior ideal f of S, for all a 2 S.
Proof. (i) ) (iii)
Let f be any fuzzy interior ideal of an intra-regular AG-groupoid S. Now
for any a 2 S, there exist x, y 2 S, such that a = (xa2 )y. Then we have
f (a) = f xa2 y f (a2 ) = f (aa) = f (a xa2 y ) = f (a((ya2 )x))

= f (a((y(aa))x)) = f (a((a(ya))x)) = f (a((x(ya))a))
= f ((x(ya))(aa)) = f ((xa)((ya)a)) f (a).
Clearly (iii) ) (ii).

(ii) ) (i)
Let J a2 be the principal two sided ideal generated by a2 . Then, CJ[a2 ]
is a fuzzy two sided ideal of S. Since a2 2 J a2 , so by (ii) we have
CJ[a2 ] (a) = CJ[a2 ] a2 = 1, hence a 2 J a2 = (Sa2 )S, which implies
that there exist x, y 2 S such that a = (xa2 )y.
Theorem 237 Let f be a fuzzy interior ideal of an intra-regular AG-

groupoid S with left identity, then f (ab) = f (ba), for all a, b in S.
Proof. Let S be an intra-regular AG-groupoid and a, b 2 S, then
2
f (ab) = f (ab) = f ((ab) (ab)) = f ((ba) (ba))
= f ((e (ba)) (ba)) = f ((ab) ((ba) e))
2
= f ((a (ba)) (be)) f (ba) = f (ba)
= f ((ba) (ba)) = f ((ab) (ab))
= f ((e (ab)) (ab)) = f ((ba) ((ab) e))
= f ((b (ab)) (ae)) f (ab):
The following three propositions are well-known.
Proposition 238 Every locally associative AG-groupoid has associative

powers.
Proposition 239 In a locally associative AG-groupoid S, am an = am+n ,

8 a 2 S and positive integers m, n.
Proposition 240 In a locally associative AG-groupoid S, (am )n = amn ,

for all a 2 S and positive integers m, n.
Theorem 241 Let f be a fuzzy semiprime interior ideal of a locally as-

sociative AG-groupoid S with left identity, then f (an ) = f (an+1 ), for all
positive integer n.
Proof. Let n be any positive integer and f be any fuzzy interior ideal of
S, then we have
2
f (an ) f (an ) = f (a2n ) f (a4n ) = f (an+2 a3n 2
)
= f ((aan+1 )a3n 2
) f (an+1 ).
An AG-groupoid S is called archimedean if for all a, b 2 S, there exist a

positive integer n such that an 2 (Sb)S.
Theorem 242 Let S be an archimedean locally associative AG-groupoid
with left identity, then every fuzzy semiprime fuzzy interior ideal of S is a
constant function.
Proof. Let f be any fuzzy semiprime fuzzy interior ideal of S and a, b 2 S.
Thus we have f (a) f (a2 ) f (a4 ) ... f (a2n ) = f (am ), where 2n = m.
Now since S is archimedean, so there exist a positive integer m and x, y 2 S
such that am = (xb)y. Therefore f (a) f ((xb)y) f (b). Similarly we can
prove that f (b) f (a). Hence f is a constant function.
An AG-groupoid S is called left (right) simple, if it contains no proper
left (right) ideal and is called simple if it contain no proper two sided ideal.
An AG-groupoid S is called fuzzy simple, if every fuzzy subset of S is a
constant function.
Theorem 243 An AG-groupoid S is simple if and only if S = a2 S =
Sa2 = (Sa2 )S, for all a in S.
Proof. It is easy.
An AG-groupoid S is called semisimple if every two-sided ideal of S is
idempotent. It is easy to prove that S is semisimple if and only if a 2
((Sa)S)((Sa)S), that is, for every a 2 S, there exist x, y, u, v 2 S such
that a = ((xa)y)((ua)v).
Theorem 244 Every fuzzy two-sided ideal of a semisimple AG-groupoid
S is an idempotent.
Proof. Let f be fuzzy
_two-sided ideal of S. Obviously
(f f )(a) = ff ((xa) y) ^ f ((ua) v)g f (a).
a=((xa)y)((ua)v)
Theorem 245 Let f and g be any fuzzy ideal of a semisimple AG-groupoid

S, then f g is a fuzzy ideal in S.
Proof. Clearly f g f \ g. Now
_
(f g) (a) = ff ((xa) y) ^ g ((ua) v)g
a=((xa)y)((ua)v)
f ((xa) y) ^ g ((ua) v) f (xa) ^ g (ua) f (a) ^ g (a)

= (f \ g) (a).
Hence f g is a fuzzy ideal of S.

Theorem 246 Every fuzzy interior ideal of a semisimple AG-groupoid S
with left identity, is a fuzzy two-sided ideal of S.
Proof. Let f be a fuzzy interior ideal of S, and a, b 2 S, then since S
is semisimple, so there exist x, y, u, v 2 S and p, q, r, s 2 S such that
a = ((xa)y)((ua)v) and b = ((pb)q)((rb)s). Thus we have
f (ab) = f (((xa)y)((ua)v)b) = f ((((xa)(ua))(yv))b)
= f ((((xu)(aa))(yv))b) = f ((((aa)(ux))(yv))b)
= f ((((yv)(ux))(aa))b) = f (((aa)((ux)(yv)))b)
= f ((b((ux)(yv)))(aa)) = f ((ba)(((ux)(yv))a))
= f (((((ux)(yv))a)a)b) f (a).
Now we have
f (ab) = f (a(((pb)q)((rb)s))) = f (a(((pb)(rb))(qs)))
= f (a(((pr)(bb))(qs))) = f (a(((bb)(rp))(qs)))
= f (a((((rp)b)b)(qs)))
= f ((((rp)b)b)(a(qs))) f (b).
Theorem 247 The set of fuzzy ideals of a semisimple AG-groupoid S

forms a semilattice structure.
Proof. Let I be the set of fuzzy ideals of a semisimple AG-groupoid S
and f , g and h 2 I , then clearly I is closed and we have f = f 2 and
f g = f \ g, where f and g are ideals of S. Clearly f g = g f , and then,
we get (f g) h = (h g) f = f (g h).
A fuzzy ideal f of an AG-groupoid S is said to be strongly irreducible if
and only if for fuzzy ideals g and h of S, g \ h f implies that g f or
h f.
The set of fuzzy ideals of an AG-groupoid S is called totally ordered
under inclusion if for any fuzzy ideals f and g of S either f g or g f .
A fuzzy ideal h of an AG-groupoid S is called fuzzy prime ideal of S, if
for any fuzzy ideals f and g of S, f g h, implies that f h or g h.
Theorem 248 In a semisimple AG-groupoid S, a fuzzy ideal is strongly
irreducible if and only if it is fuzzy prime.
Proof. It follows from theorem 245.
Theorem 249 Every fuzzy ideal of a semisimple AG-groupoid S is fuzzy
prime if and only if the set of fuzzy ideals of S is totally ordered under
inclusion.
Proof. It is easy.
7
(2; 2 _q) and (2; 2 _qk )-fuzzy
Bi-ideals of AG-groupoids
In this chapter we characterize intra-regular AG-groupoids by the proper-
ties of the lower part of (2; 2 _q)-fuzzy bi-ideals. Moreover we characterize
AG-groupoids using (2; 2 _qk )-fuzzy.
Let f be a fuzzy subset of an AG-groupoid S and t 2 (0; 1]. Then xt 2 f
means f (x) t, xt qf means f (x) + t > 1, xt _ f means xt f or xt f ,
where ; denotes any one of 2; q; 2 _q; 2 ^q. xt ^ f means xt f and
xt f , xt f means xt f does not holds.
Let f and g be any two fuzzy subsets of an AG-groupoid S, then for
k 2 [0; 1); the product f 0:5 g is de…ned by,
8 _
< ff (b) ^ g(c) ^ 0:5g , if there exist b; c 2 S, such that a = bc:
(f 0:5 g) (a) = a=bc
:
0; otherwise.
The following de…nitions for AG-groupoids are same as for semigroups
in [56].
De…nition 250 A fuzzy subset of an AG-groupoid S is called an (2
; 2 _q)-fuzzy AG-subgroupoid of S if for all x; y 2 S and t; r 2 (0; 1], it
satis…es,
xt 2 , yr 2 implies that (xy)minft;rg 2 _q .
De…nition 251 A fuzzy subset of S is called an (2; 2 _q)-fuzzy left
(right) ideal of S if for all x; y 2 S and t; r 2 (0; 1], it satis…es,
xt 2 implies (yx)t 2 _q (xt 2 implies (xy)t 2 _q ).
De…nition 252 A fuzzy AG-subgroupoid f of an AG-groupoid S is called
an (2; 2 _q)-fuzzy interior ideal of S if for all x; y; z 2 S and t; r 2 (0; 1]
the following condition holds.
yt 2 f implies ((xy)z)t 2 _qf .
De…nition 253 A fuzzy subset f of an AG-groupoid S is called an (2
; 2 _q)-fuzzy quasi-ideal of S if it satis…es, f (x) min(f CS (x); CS
f (x); 0:5), where CS is the fuzzy subset of S mapping every element of S
on 1.
De…nition 254 A fuzzy subset f of an AG-groupoid S is called an (2; 2
_q)-fuzzy generalized bi-ideal of S if xt 2 f and zr 2 f implies ((xy) z)minft;rg 2
_qf , for all x; y; z 2 S and t; r 2 (0; 1].
De…nition 255 A fuzzy subset f of an AG-groupoid S is called an (2; 2

_q)-fuzzy bi-ideal of S if for all x; y; z 2 S and t; r 2 (0; 1] the following
conditions hold
(i) If xt 2 f and yr 2 S implies (xy)minft;rg 2 _qf ,
(ii) If xt 2 f and zr 2 f implies ((xy) z)minft;rg 2 _qf .
The proofs of the following four theorems are same as in [56].
Theorem 256 Let be a fuzzy subset of S. Then is an (2; 2 _q)-fuzzy

AG-subgroupoid of S if and only if (xy) minf (x) ; (y); 0:5g.
Theorem 257 A fuzzy subset of an AG-groupoid S is called an (2; 2

_q)-fuzzy left (right) ideal of S if and only if
(xy) minf (y); 0:5g ( (xy) minf (x); 0:5g).
Theorem 258 A fuzzy subset f of an AG-groupoid S is an (2; 2 _q)-fuzzy

interior ideal of S if and only if it satis…es the following conditions.
(i) f (xy) min ff (x) ; f (y) ; 0:5g for all x; y 2 S and k 2 [0; 1).
(ii) f ((xy)z) min ff (y) ; 0:5gfor all x; y; z 2 S and k 2 [0; 1).
Theorem 259 Let f be a fuzzy subset of S. Then f is an (2; 2 _q)-fuzzy

bi-ideal of S if and only if
(i) f (xy) minff (x) ; f (y); 0:5g for all x; y 2 S and k 2 [0; 1),
(ii) f ((xy)z) minff (x); f (z) ; 0:5g for all x; y; z 2 S and k 2 [0; 1).
Example 260 Let S = fa; b; cg be an AG-groupoid with the following Cay-

ley table:
a b c
a a a a
b a a a
c a c c
One can easily check that fag; fbg; fcg; fa; cg and fa; b; cg are all bi-ideals
of S. Let f be fuzzy subsets of S such that f (a) = 0:9; f (b) = 0:2 and
f (c) = 0:6: Then f is an (2; 2 _q)-fuzzy ideal of S:
De…nition 261 An element a of an AG-groupoid S is called intra-regular

if there exist x; y in S such that a = (xa2 )y. An AG-groupoid S is called
intra-regular if every element of S is intra-regular.
Theorem 262 Let S be an AG-groupoid with left identity. Then S is intra-

regular if and only if R \ L = RL and R is semiprime, for every left ideal
L and every right ideal R of S.
Proof. Let R, L be right and left ideals of an intra-regular AG-groupoid

S with left identity. Then R and L become ideals of S and so RL R \ L.
Now we have R \ L is an ideal of S. We can also deduce that R \ L =
(R \ L)2 RL. Hence we obtain R \ L = RL.
Next, we show that R is semiprime. So let r2 2 R. Since S is intra-regular,

there exist x; y 2 S such that r = (xr2 )y. Thus we have
r = (xr2 )y 2 (SR)S R:
Therefore, R is semiprime.
Conversely, assume that R \ L = RL and R is semiprime, for any left
ideal L and right ideal R of S. We need to show that S is intra-regular. To
see this, note that for any a 2 S, Sa2 is a right ideal and Sa is a left ideal
of S. Clearly, a 2 Sa. Since Sa2 is semiprime and a2 2 Sa2 , we also have
a 2 Sa2 . Hence it follows that
a 2 Sa2 \ Sa = (Sa2 )(Sa) (Sa2 )S;
which shows that a is intra-regular. Therefore, S is intra-regular as required.
Example 263 Let S = f1; 2; 3; 4; 5g be an AG-groupoid with the following

Cayley table:
1 2 3 4 5
1 4 5 1 2 3
2 3 4 5 1 2
3 2 3 4 5 1
4 1 2 3 4 5
5 5 1 2 3 4
It is clear that S is intra-regular since 1 = (3 12 ) 2; 2 = (1 22 ) 5;

3 = (5 32 ) 2; 4 = (2 42 ) 1; 5 = (3 52 ) 1. Let us de…ne a fuzzy
subset f on S such that f (1) = 0:8; f (2) = 0:7; f (3) = 0:5; f (4) = 0:9
and f (5) = 0:6. Then f is an (2; 2 _q)-fuzzy bi-ideal of S.
Theorem 264 For an AG-groupoid S with left identity, the following con-
ditions are equivalent:
(ii) B = B 2 for every bi-ideal B of S.
Proof. Let B be a bi-ideal of an intra-regular AG-groupoid S with left
identity. Thus B is an ideal of S. Then it follows that B = B 2 .
Conversely, assume that B = B 2 for every bi-ideal B of S. For any a 2 S,
Sa is a bi-ideal contains a. Thus we have
a 2 Sa = (Sa)2 = Sa2 = (Sa2 )S;
which shows that a is intra-regular. Therefore, S is intra-regular as required.
The following results can be proved by similar techniques as used in [56].

Lemma 265 Let L be a non-empty subset of an AG-groupoid S and CL

be the characteristic function of L. Then L is a left ideal of S if and only
if the lower part CL is an (2; 2 _q)-fuzzy left ideal of S.
Lemma 266 Let R be a non-empty subset of an AG-groupoid S and CR

be the characteristic function of R. Then R is a right ideal of S if and only
if the lower part CR is an (2; 2 _q)-fuzzy right ideal of S.
Lemma 267 Let A and B be non-empty subsets of an AG-groupoid S.

Then we have the following:
(1) (CA ^ CB ) = C(A\B) .
(2) (CA _ CB ) = C(A[B) .
(3) (CA CB ) = C(AB) .
Theorem 268 A fuzzy subset f of an AG-groupoid S is (2; 2 _q)-fuzzy

semiprime if and only if f (x) f (x2 ) ^ 0:5 for all x 2 S.
Proof. Let f be a fuzzy subset of an AG-groupoid S which is (2; 2 _q)-

fuzzy semiprime. If there exists some x0 2 S such that f (x0 ) < t0 =
f x20 ^ 0:5. Then (x20 )t0 2 f , but (x0 )t0 2f . In addition, we have (x0 )t0 2
_qf since f is (2; 2 _q)-fuzzy semiprime. On the other hand, we have
f (x0 ) + t0 t0 + t0 1. Thus (x0 )t0 qf , and so (x0 )t0 2 _qf . This is a
contradiction. Hence f (x) f (x2 ) ^ 0:5 for all x 2 S.
Conversely, assume that f is a fuzzy subset of an AG-groupoid S such
that f (x) f (x2 ) ^ 0:5 for all x 2 S. Let x2t 2 f: Then f x2 t, and so
f (x) f x2 ^ 0:5 t ^ 0:5. Now, we consider the following two cases:
(i) If t 0:5, then f (x) t. That is, xt 2 f . Thus we have xt 2 _qf:
(ii) If t > 0:5, then f (x) 0:5. It follows that f (x) + t 0:5 + t > 1.
That is, xt qf , and so xt 2 _qf also holds. Therefore, we conclude that f
is (2; 2 _q)-fuzzy semiprime as required.
Theorem 269 Let A be a non-empty subset of an AG-groupoid S with left

identity. Then A is semiprime if and only if CA is fuzzy semiprime.
Proof. Suppose that A is a non-empty subset of an AG-groupoid S with

left identity and A is semiprime. For any a 2 S, if a2 2 A, then we have
a 2 A since A is semiprime. It follows that CA (a) = CA (a2 ) = 0:5. If
a2 2A, then we have CA (a) 0 = CA (a2 ). This shows that CA is fuzzy
semiprime.
Conversely, assume that CA is fuzzy semiprime. Thus we have CA (x)
CA (x2 ) for all x in S. If x2 2 A, then CA (x2 ) = 0:5. Hence CA (x)
CA (x2 ) 0:5, which implies x 2 A. Therefore, A is semiprime as required.
De…nition 270 An (2; 2 _q)-fuzzy AG-subgroupoid of an AG-groupoid S

is called an (2; 2 _q)-fuzzy interior ideal of S if
at 2 f ) ((xa)y)t 2 _qf;
for all x; a; y 2 S and t 2 (0; 1].
Theorem 271 Let f be a fuzzy subset of an AG-groupoid S. Then f is an

(2; 2 _q)-fuzzy interior ideal of S if and only if it satis…es:
(i) f (xy) minff (x); f (y); 0:5g; for all x; y 2 S:
(ii) f ((xa)y) minff (a); 0:5g; for all x; a; y 2 S.
Proof. It is easy.

AG-groupoids
In this section, we give some characterizations of intra-regular AG-groupoids
based on the properties of their (2; 2 _q)-fuzzy ideals.
Theorem 272 For an AG-groupoid S with left identity e, the following

conditions are equivalent:
(ii) (f ^g) = (f g) and f is fuzzy semiprime, where f is an (2; 2 _q)-
fuzzy right ideal and g is an (2; 2 _q)-fuzzy left ideal of S.
Proof. Assume that S is an intra-regular AG-groupoid with left identity

e. Let f be an (2; 2 _q)-fuzzy right ideal of S and g be (2; 2 _q)-left ideal
of S. For a 2 S, we have
_
(f g) (a) = (f ^ g)(a) ^ 0:5 = ( ff (y) ^ g(z)g) ^ 0:5
a=yz
_
=( (ff (y) ^ g(z)g ^ 0:5)
a=yz
_
=( (ff (y) ^ 0:5g ^ fg(z) ^ 0:5g ^ 0:5)
a=yz
_
ff (yz) ^ g(yz) ^ 0:5g
a=yz
= f (a) ^ g(a) ^ 0:5 = (f ^ g) (a):

Thus (f g) (f ^ g) . Since S is intra-regular, for a 2 S, there exist

x; y 2 S such that a = (xa2 )y: Now we get
a = (xa2 )y = (x(a:a)y = (a(xa))y = (y(xa))a
= ((ey)(xa))a = ((ax) (ye)) = ((ay)(xe))a
= ((ax)(ye))a = ((ax)(ye))((xa2 )y) = ((ax)(ye))((x(aa))y)
= ((ax)(ye))((a(xa))y) = ((ax)(ye))((y(xa))a):
Hence we deduce that
(f g) (a) = (f g)(a) ^ 0:5
_
= ff (p) ^ g(q)g ^ 0:5
a=pq
_
=[ ff (p) ^ g(q)g] ^ 0:5
a=pq
_
= ff (p) ^ g(q)g ^ 0:5
pq=((ax)(ye))((y(xa))a)
f ((ax)(ye)) ^ g((y(xa))a) ^ 0:5

f (ax) ^ g(a) ^ 0:5 f (a) ^ g(a) ^ 0:5
= (f ^ g)(a) ^ 0:5 = (f ^ g )(a):
This shows that (f g) (f ^ g) : Thus we obtain (f g) = (f ^ g) .
Next we shall show that f is fuzzy semiprime. Since S = S 2 ; thus for
x 2 S there exist u; v in S such that x = uv: Then we get
f (a) = f xa2 y f xa2 = f ((uv) (aa)) = f ((aa) (vu)) f a2 :
Therefore, f is fuzzy semiprime as required.
Conversely, suppose that S is an AG-groupoid with left identity e, such
that (f ^ g) = (f g) and f is fuzzy semiprime for every (2; 2 _q)-fuzzy
right ideal f and every (2; 2 _q)-fuzzy left ideal g of S. Let R and L be
right and left ideals of S respectively. Then, CL and CR are (2; 2 _q)-fuzzy
left ideal and (2; 2 _q)-fuzzy right ideal of S; respectively. By assumption,
CR is also fuzzy semiprime. Then we deduce that R is semiprime. Then we
have
C(RL) = (CR CL ) = (CR ^ CL ) = C(R\L) .
Thus RL = R \ L. Hence S is intra-regular as required.
Note that RL R \ L for every right ideal R and left ideal L of an
AG-groupoid S. We immediately obtain the following.
Theorem 273 For an AG-groupoid S with left identity e, the following
(ii) (f ^g) (f g) and f is fuzzy semiprime, where f is an (2; 2 _q)-
fuzzy right ideal and g is an (2; 2 _q)-fuzzy left ideal of S.
Theorem 274 For an AG-groupoid S with left identity e; the following

(ii) ((h ^ f ) ^ g) ((h f ) g) and h is fuzzy semiprime, for every
(2; 2 _q)-fuzzy right ideal h, (2; 2 _q)-fuzzy bi-ideal f and (2; 2 _q)-fuzzy
left ideal g of S.
(iii) ((h ^ f ) ^ g) ((h f ) g) and h is fuzzy semiprime, for every
(2; 2 _q)-fuzzy right ideal h, (2; 2 _q)-fuzzy generalized bi-ideal f and
(2; 2 _q)-fuzzy left ideal g of S.
Proof. (i) ) (iii): Let S be an intra-regular AG-groupoid with left identity
e. For any a 2 S, there exist x and y in S such that a = (xa2 )y. Now we
get

= ((ey)(xa))a = ((ye)(xa))a = (ye) x xa2 y a
2 2
= ((ye)(x((xa )(ey)))a = ((ye)(x((ye)(a x))a
= ((ye)(x(a2 ((ye)x))a = (((ye) (a2 (x((ye)x))))a
= ((ye)(a2 ((ye)x2 )))a = (a2 ((ye)((ye)x2 )))a
= ((ba)a)a, where b = (ye) ((ye)x2 ):
Now, we have
ba = b((xa2 )y) = (xa2 )(by) = (yb)(xa2 )

= (yx)(a2 b) = a2 (yx)b = (((yx)b)a)a
= (((yx)b)((xa2 )y))a = (t((xa2 )y))a; where t = (yx)b:
(t((xa2 )y))a = (((xa2 )(ty)))a = (((yt)(a2 x)))a
= (a2 ((yt)x))a = (a2 u)a; where u = (yt)x:
Thus a = (((a2 u)a)a)a: Furthermore, we can deduce that

_
((h f ) g )(a) = [ ((h f )(p) ^ g(q) ^ 0:5]
a=pq
_
= [(h f )(p) ^ g(q)] ^ 0:5
pq=((ba)a)a
_
= [(h f )((ba)a) ^ g(a)] ^ 0:5
pq=((ba)a)a
(h f )((ba)a) ^ g(a) ^ 0:5

W
= fh((a2 u)a) ^ f (a) ^ 0:5g ^ g(a) ^ 0:5
(ba)a=((a2 u)a)a
h(a2 ) ^ f (a) ^ g(a) ^ 0:5

h(a) ^ f (a) ^ g(a) ^ 0:5 = ((h ^ f ) ^ g )(a).
This shows that ((h ^ f ) ^ g ) ((h f ) g ). In addition, for any a 2 S;

there exist x; y 2 S such that a = xa2 y since S is intra-regular. Thus we
get
h (a) = h xa2 y = h(xa2 ) = h(x(aa)) = h(a(xa)) = h((ea) (xa))

= h (ax) (ae) = h(aa)(xe) = h a2 (xe) h a2 :
Therefore, h is fuzzy semiprime as required.

(iii) ) (ii): Straightforward.
(ii) ) (i): Let h be an (2; 2 _q)-fuzzy semiprime right ideal and let g be
an (2; 2 _q)-fuzzy left ideal of S. Then
_ _
(h Cs )(a) = h(b) ^ Cs (c) = h(b) ^ 1
a=bc a=bc
_
h(bc) = h(a):
a=bc
Since Cs is an (2; 2 _q)-fuzzy bi-ideal of S, for any a 2 S we have
(h ^ g )(a) = ((h ^ Cs ) ^ g )(a) ((h Cs ) g )(a) (h g )(a).
Therefore, (h ^ g ) (h g ). Then by Theorem 273, we deduce that S is

intra-regular as required.
Lemma 275 A non-empty subset B of an AG-groupoid S is a bi-ideal if

and only if CB is an (2; 2 _q)-fuzzy bi-ideal of S:
Proof. It is similar to the proof of Lemma 9 in [55].

(ii) f = ((f Cs ) f ) for every (2; 2 _q)-fuzzy generalized bi-ideal f
of S and f f = f .
(iii) f = ((f Cs ) f ) for every (2; 2 _q)-fuzzy bi-ideal f of S and
f f = f.
Proof. (i) ) (ii): Let S be an intra-regular AG-groupoid with left identity

e. For any a 2 S, there exist x and y in S such that a = (xa2 )y. We already
obtained a = (((a2 u)a)a)a: Moreover, we have
((f CS ) f ) (a) = ((f CS ) f )(a) ^ 0:5

_
= f(f CS )(p) ^ f (q)g ^ 0:5
a=pq
_
= f(f CS )(p) ^ f (q)g ^ 0:5
pq=(((a2 u)a)a)a
(f CS )(((a2 u)a)a) ^ f (a) ^ 0:5

_
= ff ((a2 u)a) ^ CS (a)g ^ f (a) ^ 0:5
bc=((a2 u)a)a
f ((a2 u)a) ^ 1 ^ f (a) ^ 0:5

f (a2 ) ^ f (a) ^ 0:5 ^ f (a) ^ 0:5
f (a) ^ f (a) ^ 0:5 f (a) ^ 0:5 = f (a);
which shows that ((f CS ) f ) f :

On the other hand, since f is an (2; 2 _q)-fuzzy generalized bi-ideal of
S, we can deduce that
((f CS ) f ) (a) = ((f CS ) f )(a) ^ 0:5

_
= f(f CS )(b) ^ f (c)g ^ 0:5
a=bc
_ W
= f ff (p) ^ CS (q)g ^ f (c)g ^ 0:5
a=bc b=pq
_ W
= f ff (p) ^ 1g ^ f (c)g ^ 0:5
a=yz b=pq
_ W
= f ff (p)g ^ f (c)g ^ 0:5
a=bc b=pq
W
ff ((pq)c) ^ 0:5g f (a) ^ 0:5 = f (a):
a=(pq)c
Thus ((f CS ) f ) _f , and so ((f CS ) f ) = f as required.

Now (CS f )(a) = fCS (e)^f (a)g = f (a): Therefore, f f CS f =
a=ea
f: Since S is intra-regular, thus we get
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a:
In addition, we also have
y(xa) = y(x(xa2 )y)) = y((xa2 )(xy)) = (xa2 )(y(xy))

= (xa2 )(xy 2 ) = (a(xa))(xy 2 ) = ((xy 2 )(xa))a:
Let us denote xy 2 = t. Then we deduce that
((xy 2 )(xa) = t(xa) = t(x((xa2 )y)) = t((xa2 )(xy))

= (xa2 )(t(xy)) = ((xy)t)(a2 x) = a2 (((xy)t)x)
= (x(xy)t))(aa) = a((x((xy)t))a)
Thus we obtain a = (y(xa))a = (a((x((xy)t))a)a)a. Now, it follows that

_
f f (a) = ff (y(xa)) ^ f (a)g
a=(y(xa))a
_
= ff (a((x((xy)t))a)) ^ f (a)g
y(xa)=a((x((xy)t))a)a
f (a) ^ f (a) f (a);
which shows that f f f: Therefore, f f = f:

(ii) ) (iii): Straightforward.
(iii) ) (i): Let B be any bi-ideal of an AG-groupoid S with left identity
e. Then CB is an (2; 2 _q)-fuzzy bi-ideal of S. Thus we have CB CB = CB .
Also it is clear that CB CB = CB 2 . Hence CB = CB 2 and so B = B 2 .
Hence deduce that S is intra-regular as required.

(ii) (f ^ g) = ((f g) f ) for every (2; 2 _q)-fuzzy bi-ideal f and
(2; 2 _q)-fuzzy interior ideal g of S:
Proof. (i) ) (ii): Let S be an intra-regular AG-groupoid with left identity
e. Let f be an (2; 2 _q)-fuzzy bi-ideal and g be an (2; 2 _q)-fuzzy interior
ideal of S. Since Cs itself is an (2; 2 _q)-fuzzy ideal of S, for any a 2 S,
we have
((f g) f ) (a) ((f Cs ) f )(a) ^ 0:5

_
= f(f Cs )(p) ^ f (q)g ^ 0:5
a=pq
_ _
= f ff (b) ^ Cs (c)g ^ f (q)g ^ 0:5
a=pq p=bc
_ _
= f ff (b) ^ 1g ^ f (q)g ^ 0:5
a=pq p=bc
_ _
= f f (b) ^ f (q)g ^ 0:5
a=pq p=bc
_
ff ((bc)qg ^ 0:5 = f (a) ^ 0:5 = f (a):
a=(bc)q
Note also that
((f g) f ) (a) ((Cs g) Cs )(a) ^ 0:5

_
= f(Cs g)(b) ^ Cs (c)g ^ 0:5
a=bc
_ _
= f fCs (p) ^ g(q)g ^ 1g ^ 0:5
a=bc b=pq
_ _
= f1 ^ g(q)g ^ 0:5 = fg(q)g ^ 0:5
a=(pq)c) a=(pq)c)
_
fg((pq)c)g ^ 0:5 = g(a) ^ 0:5 = g (a).
a=(pq)c)
Hence ((f g) f ) (f ^ g ) = (f ^ g) . Now, since S is intra-regular,

for a 2 S there exist elements x; y 2 S such that a = (xa2 )y. We already
obtained a = (((a2 u)a)a)a: Thus we have
_
((f g) f ) (a) = f(f g)(y(xa)) ^ f (a)g ^ 0:5
a=(((a2 u)a)a)a
(f g)(((a2 u)a)a) ^ f (a) ^ 0:5

_
= ff (((a2 u)a) ^ g(a) ^ 0:5g ^ f (a) ^ 0:5
((a2 u)a)a=bc
_
= f (((a2 u)a) ^ g(a) ^ f (a) ^ 0:5
((a2 u)a)a=bc
f (a2 ) ^ f (a) ^ 0:5 ^ g(a) ^ f (a) ^ 0:5

f (a) ^ f (a) ^ g(a) ^ f (a) ^ 0:5
f (a) ^ g(a) ^ f (a) ^ 0:5
f (a) ^ g(a) ^ 0:5 = (f ^ g) (a);
which gives ((f g) f ) (f ^ g) . Therefore, ((f g) f ) = (f ^ g)

as required.
(ii) ) (i): Assume that S is an AG-groupoid with left identity such that
(f ^ g) = ((f g) f ) for every (2; 2 _q)-fuzzy bi-ideal f and (2; 2 _q)-
fuzzy interior ideal g of S: Let f be any (2; 2 _q)-fuzzy bi-ideal of S. Since
CS itself is an (2; 2 _q)-fuzzy interior ideal of S, we have
f (a) = f (a) ^ 0:5 = (f ^ CS )(a) ^ 0:5 = (f ^ CS ) (a) = ((f CS ) f ) (a);
for all a 2 S. That is, ((f CS ) f) = f . Hence S is intra-regular as

required.
Theorem 278 For AG-groupoid S with left identity e; the following con-
ditions are equivalent:
(ii) A \ B AB, for every bi-ideal B and left ideal A of S.
(iii) (f ^ g) (f g) for every (2; 2 _q)-fuzzy bi-ideal f and (2; 2 _q
)-fuzzy left ideal g of S.
(iv) (f ^ g) (f g) for every (2; 2 _q)-fuzzy generalized bi-ideal f
and every (2; 2 _q)-fuzzy left ideal g of S.
Proof. (i) ) (iv): Let S be an intra-regular AG-groupoid with left identity

e. Let f and g be any (2; 2 _q)-fuzzy generalized bi-ideal and any (2;
2 _q)-fuzzy left ideal of S; respectively. For any a 2 S, there exist x and y
in S such that a = (xa2 )y. Thus
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = ((ey)(xa))a

= ((ax)(ye))a = (((ye)x)a)a = (ta)a; where t = (ye)x:
Thus
a = (ta)a = (t(xa2 )y)a = ((xa2 )(ty))a = ((x(aa))(ty))a = ((a(xa))(ty))a:
Furthermore, we have
(f g) (a) = (f g)(a) ^ 0:5

_
= ff (p) ^ g(q)g ^ 0:5
a=pq
_
= ff (p) ^ g(q)g ^ 0:5
a=((a(xa))(ty))a
f (a(xa)) ^ g((ty)a) ^ 0:5

f (a) ^ f (a) ^ g(a) ^ 0:5
= f (a) ^ g(a) ^ 0:5
= (f ^ g) ^ 0:5 = (f ^ g) (a):
That is, (f ^ g) (f g) .
(iv) ) (iii): Straightforward.
(iii) ) (ii): Assume that S is an AG-groupoid with left identity such that
(f ^g) (f g) for every (2; 2 _q)-fuzzy bi-ideal f and (2; 2 _q )-fuzzy
left ideal g of S. Let A and B be bi-ideal and left ideal of S, respectively.
Then CA and CB are (2; 2 _q)-fuzzy bi-ideal and (2; 2 _q)-fuzzy left ideal
of S. Thus by hypothesis we get
CA\B = (CA ^ CB ) (CA CB ) = CAB .
It follows that A \ B AB.

(ii) ) (i): Since Sa is both a bi-ideal and left ideal of an AG-groupoid
S with left identity. Using the medial law, the left invertive law and the
paramedial law, we have
a 2 Sa \ Sa = (Sa)(Sa) = (SS)(aa) = (a2 S)S

= ((aa)(SS))S = ((SS)(aa))S = (Sa2 )S;
for all a 2 S. Hence S is intra-regular as required.

(ii) (f f ) f for every (2; 2 _q)-fuzzy bi-ideal f of S:
(iii) (f g) f ^ g for every (2; 2 _q)-fuzzy bi-ideals f and g of S:
Proof. (i) ) (iii): Let S be an intra-regular AG-groupoid with left identity
e. Let f and g be (2; 2 _q)-fuzzy bi-ideals of S: For any a 2 S, there exist
x; y in S such that a = xa2 y: Thus we get
a = xa2 y = (x (aa)) y = (a (xa)) y = (y(xa))a;
and
y(xa) = y(x (xa2 )y ) = y (xa2 (xy))

= xa2 xy 2 = (a (xa)) xy 2 = ((xy 2 )(xa))a
= ( xy 2 (x((xa2 )y))a = ((xy 2 )((xa2 )(xy)))a
= ((xa2 )((xy 2 ))(a2 x))a = ( (xy) xy 2 (a2 x))a
= (a2 (((xy) (xy 2 ))x))a = ((x((xy)(xy 2 )))(aa))a
= a x((xy) xy 2 )a a:
Thus a = (y(xa))a = a x((xy) xy 2 )a a: Now, we have
(f g) (a) = (f g) (a) ^ 0:5

0 1
_
=@ f a x((xy) xy 2 )a ^ g (a)A ^ 0:5
a=(a(x((xy)(xy 2 )))a)a
0 1
_
=@ f a x((xy) xy 2 )a ^ g (a)A ^ 0:5
a=(a(x((xy)(xy 2 )))a)a
f a x((xy) xy 2 )a ^ g (a) ^ 0:5

(f (a) ^ f (a) ^ 0:5) ^ (g (a) ^ 0:5)
[f (a) ^ g (a)] ^ 0:5 = (f ^ g) (a) ^ 0:5 = (f ^ g) (a) :
This shows that (f g) f ^g as required.

(iii) ) (ii): Straightforward.
(ii) ) (i): Assume that S is an AG-groupoid with left identity e such

that (f f ) f for every (2; 2 _q)-fuzzy bi-ideal f of S. Let B be a bi-
ideal of S. Then CB is an (2; 2 _q)-fuzzy bi-ideal of S: By hypothesis, we
have (CB CB ) = CB 2 CB , and so B B 2 . Clearly, we have B 2 B
since B is a bi-ideal of S. Therefore, B 2 = B. Hence S is intra-regular.
7.2 (2; 2 _qk )-fuzzy Ideals of Abel-Grassmann’s

7.3 Main results
We begin with the following de…nition.
De…nition 280 An element a of an AG-groupoid S is called intra-regular

if there exists x; y 2 S such that a = (xa2 )y and S is called intra-regular,
if every element of S is intra-regular.
Let S be an intra-regular AG-groupoid with left identity. Then, for x

in S there exist u and v in S such that x = uv. Now, using paramedial,
medial, left invertive law, we get
a = (xa2 )y = [(uv)(aa)]y = [(aa)(vu)]y = [y(vu)]a2 = a[fy(vu)ga](2)

= [y(vu)]a2 = (ya)[(vu)a] = [a(vu)](ay) = [(ay)(vu)]a.
Note. It is obvious from (2) that the results for intra-regular AG-groupoid
with left identity is signi…cantly di¤erent from those of semigroups and
monoids.
The characteristic function CA for a subset A of an AG-groupoid S is
de…ned by
1, if x 2 A;
CA (x) =
0, if x 2
= A:
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy subgroupoid of S if
for all x; y 2 S and t; r 2 (0; 1] the following condition holds.
xt 2 f and yr 2 f implies that (xy)minft;rg 2 _qk f:
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy left(right) ideal of S
if for all x; y 2 S and t; r 2 (0; 1] the following condition holds.
yt 2 f implies that (xy)t 2 _qk f (yt 2 f implies that (yx)t 2 _qk f ) :
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy two sided ideal of S
if it is both (2; 2 _qk )-fuzzy left and (2; 2 _qk )-fuzzy right ideal of S.
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy bi-ideal of S if for all
x; y; z 2 S and t; r 2 (0; 1] the following condition holds.
xt 2 f and zr 2 f implies that ((xy) z)minft;rg 2 _qk f:
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy generalized bi-ideal

of S if for all x; y; z 2 S and t; r 2 (0; 1] the following condition holds.
xt 2 f and zr 2 f implies that ((xy) z)minft;rg 2 _qk f:
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy interior of S if for all
x; y; z 2 S and t; r 2 (0; 1] the following condition holds.
(a) xt 2 f and yr 2 f implies that (xy)minft;rg 2 _qk f:
(b) at 2 f implies that ((xa) y)t 2 _qk f:
A fuzzy subset f of S is called an (2; 2 _qk )-fuzzy generalized interior
of S if for all x; y; z 2 S and t 2 (0; 1] the following condition holds.
at 2 f implies that ((xa) y)t 2 _qk f:
A fuzzy subset f of an AG-groupoid S is called (2; 2 _qk )-fuzzy semi-
prime if f (a) f (a2 ) ^ 1 2 k , for all a in S.
De…nition 281 Let A be any subset of S. Then, the characteristic function
(CA )k is de…ned as,
1 k
(CA )k (x) = 2 , if x 2 A;
0, if x 2= A:
The proof of the following two lemma’s are same as in [55].

Lemma 282 For an AG-groupoid S, the following holds.
(i) A non empty subset J of AG-groupoid S is an ideal if and only if
(CJ )k is an (2; 2 _qk )-fuzzy ideal.
(ii) A non empty subset L of AG-groupoid S is left ideal if and only if
(CL )k is an (2; 2 _qk )-fuzzy left ideal.
(iii) A non empty subset R of AG-groupoid S is right ideal if and only
if (CR )k is an (2; 2 _qk )-fuzzy right ideal.
(iv) A non empty subset B of AG-groupoid S is an bi-ideal if and only
if (CB )k is an (2; 2 _qk )-fuzzy bi-ideal.
(v) A non empty subset I of AG-groupoid S is an interior ideal if and
only if (CI )k is an (2; 2 _qk )-fuzzy interior ideal.
(vi) A non empty subset I of AG-groupoid S is semiprime if and only if
(CI )k is an (2; 2 _qk )-fuzzy semiprime.
(vii) A right ideal R of an AG-groupoid S is semiprime if and only if
(CR )k is (2; 2 _qk )-fuzzy semiprime.
Let f and g be any two fuzzy subsets of an AG-groupoid S. Then, for
k 2 [0; 1); the product f k g is de…ned by,
8 _
< 1 k
f (b) ^ g(c) ^ , if there exist b; c 2 S, such that a = bc:
(f k g) (a) = 2
: a=bc
0; otherwise.
De…nition 283 Let f and g be fuzzy subsets of an AG-groupoid S. We

de…ne the fuzzy subsets fk , f ^k g, f _k g and f k g of S as follows,
(i) fk (a) = f (a) ^ 1 2 k .

(ii) (f ^k g) (a) = (f ^ g) (a) ^ 1 2 k .
(iii) (f _k g) (a) = (f _ g) (a) ^ 1 2 k .
(iv) (f k g) (a) = (f g) (a) ^ 1 2 k , for all a 2 S.
Lemma 284 Let A; B be non empty subsets of an AG-groupoid S. Then,

the following holds.
(i) (CA\B )k = (CA ^k CB ) :
(ii) (CA[B )k = (CA _k CB ) :
(iii) (CAB )k = (CA k CB ) :
Example 285 Let S = f1; 2; 3; 4; 5; 6g, and the binary operation " " be
de…ned on S as follows.
1 2 3 4 5 6
1 1 1 1 1 1 1
2 1 2 1 1 1 1
3 1 1 5 6 3 4
4 1 1 4 5 6 3
5 1 1 3 4 5 6
6 1 1 6 3 4 5
Clearly 1 = (1 12 ) 1; 2 = (2 22 ) 2; = 3(3 32 ) 5; 4 = (6 42 ) 3; 5 =
(5 52 ) 5; 6 = (4 62 ) 3. Clearly f1g ; f1; 2g; f1; 2; 3g ; f1; 2; 3; 4g ; f1; 2; 3; 4; 5; g
and f1; 2; 3; 4; 5; 6g are ideal of S. De…ne a fuzzy subset f : S ! [0; 1] as
follows: 8
>
> 0:9 for x = 1
>
>
>
> 0:8 for x = 2
<
0:5 for x = 3
f (x) =
>
> 0:5 for x = 4
>
>
>
> 0:5 for x = 5
:
0:5 for x = 6
Then, clearly f is an (2; 2 _qk )-fuzzy ideal of S.
equivalent.
(i) S is intra-regular. (ii) For bi-ideals B1 and B2 of S, B1 \ B2 =
(B1 B2 )B1 . (iii) For (2; 2 _qk )-fuzzy bi-ideals f and g of S; f ^k g
(f k g) k f . (iv) For (2; 2 _qk )-fuzzy generalized bi-ideals f and g of
S; f ^k g (f k g) k f .
Proof. (i) =) (iv) : Let f and g be (2; 2 _qk )-fuzzy generalized bi-ideals
of an intra-regular AG-groupoid S. Since S is intra-regular therefore for
a 2 S there exists x; y 2 S such that a = xa2 y. Now, by using left
invertive law, medial law and paramedial law we get,
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = y(x((xa2 )y)) a

= y((xa2 )(xy)) a = (xa2 )(y(xy)) a = (xa2 )(x(yy)) a
= (xa2 )(xy 2 ) a = (xx)(a2 y 2 ) a = x2 (a2 y 2 ) a
= a2 (x2 y 2 ) a = (aa)(x2 y 2 ) a = (y 2 a)(x2 a) a
= (y 2 x2 )(aa) a = a((y 2 x2 )a) a = a((y 2 x2 )((xa2 )y)) a
= a((xa2 )((y 2 x2 )y)) a = a((x(y 2 x2 ))(a2 y)) a
= a(a2 ((x(y 2 x2 ))y)) a = a2 (a((x(y 2 x2 ))y)) a
= (aa)(a((x(y 2 x2 ))y)) a = ((a((x(y 2 x2 ))y))a)a a:
Thus,
((f k g) k f )(a)
_ 1 k
= (f k g)(p) ^ f (q) ^
a=pq
2
( ) !
_ _ 1 k 1 k
= f (u) ^ g(v) ^ ^ f (q) ^
a=pq p=uv
2 2
_ 1 k
= ff (u) ^ g(v)g ^ f (q) ^
2
a=(uv)q
_ 1 k
= ff (u) ^ g(v)g ^ f (q) ^
2
a=(((a((x(y 2 x2 ))y))a)a)a=(uv)q
1 k
f (a((x(y 2 x2 ))y)))a) ^ g (a) ^ f (a) ^
2
1 k 1 k
f (a) ^ ^ g(a) ^ f (a) ^
2 2
1 k
= ff (a) ^ g(a)g ^ f (a) ^
2
1 k
= f (a) ^ g(a) ^ f (a) ^
2
1 k
= (f ^ g) (a) ^ = (f ^k g) (a) :
2
So, f ^k g (f k g) k f .
(iv) =) (iii) : is obvious.
(iii) =) (ii) : Assume that B1 and B2 are bi-ideals of S. Then (CB1 )k
and (CB2 )k are (2; 2 _qk )-fuzzy bi-ideals. Thus we have, (CB1 \B2 )k =
(CB1 ^k CB2 ) (CB1 k CB2 ) k CB1 = (C(B1 B2 )B1 )k : Hence, B1 \ B2
(B1 B2 )B1 :
(ii) =) (i) : Since Sa is a bi-ideal of S contains a. Thus, using (ii), and

medial law we have,
a 2 Sa \ Sa ((Sa)(Sa))Sa = ((SS)(aa))(SS) = (Sa2 )S:
Hence, S is intra-regular.
Theorem 287 For an AG-groupoid S with left identity; the following are
equivalent.
(i) S is intra-regular. (ii) For left ideals, L1 ; L2 of S; L1 \ L2 L1 L2 \
L2 L1 : (iii) For (2; 2 _qk )-fuzzy left ideals, f; g of S; f ^k g f k g ^g k f .
Proof. (i) =) (iii) : Let f and g be (2; 2 _qk )-fuzzy left ideals of an intra-
regular AG-groupoid S respectively. Since S is intra-regular therefore for
a 2 S there exists x; y 2 S such that a = xa2 y. Now, using left invertive
law we get,
a = xa2 y = (x (aa)) y = (a (xa)) y = (y (xa)) a. Therefore,
_ 1 k
(f k g) (a) = f (p) ^ g (q) ^
a=pq
2
_ 1 k
= f (p) ^ g (q) ^
2
a=(y(xa))a=pq
1 k
f (y (xa)) ^ g (a) ^
2
1 k 1 k
f (a) ^ ^ g (a) ^
2 2
1 k
= f (a) ^ g (a) ^ = (f ^k g) (a) :
2
Thus, f ^k g f k g. Similarly, we can show that f ^k g g k f . Thus,
we have f ^k g f k g \ g k f:
(iii) =) (ii) : Assume that L1 and L2 be any left ideals of S. Then,
(CL1 )k and (CL2 )k are (2; 2 _qk )-fuzzy left ideals of S therefore, we have,
(CL1 \L2 )k = (CL1 ^k CL2 ) (CL1 k CL2 ) = (CL1 L2 )k :
This implies that L1 \ L2 L1 L2 . Similarly, we can show that L1 \ L2

L2 L1 . Thus, L1 \ L2 L1 L2 \ L2 L1 :
(ii) =) (i) : Since Sa is a left ideal of S contains a. Thus, using (ii),
paramedial law, medial law, we get,
a 2 Sa \ Sa (Sa)(Sa) \ (Sa)(Sa) (Sa)(Sa) = (aa)(SS)

= S((aa)S) = (SS)((aa)S) = (Sa2 )SS = (Sa2 )S.
equivalent.
(i) S is intra-regular. (ii) For bi-ideal B, right ideal R ,and left ideal
L of S; B \ R \ L (BR)L and R is semiprime. (iii) For (2; 2 _qk )-fuzzy
bi-ideal f , (2; 2 _qk )-fuzzy right ideal g, and (2; 2 _qk )-fuzzy left ideal h of
S; (f ^k g) ^k h (f k g) k h and g is (2; 2 _qk )-fuzzy semiprime. (iv)
For (2; 2 _qk )-fuzzy bi-ideal f , (2; 2 _qk )-fuzzy interior ideal g, and (2; 2
_qk )-fuzzy left ideal h of S; (f ^k g) ^k h (f k g) k h and g is (2; 2 _qk )-
fuzzy semiprime. (v) For (2; 2 _qk )-fuzzy generalized bi-ideal f , (2; 2 _qk )-
fuzzy generalized interior ideal g, and (2; 2 _qk )-fuzzy left ideal h of S;
(f ^k g) ^k h (f k g) k h and g is (2; 2 _qk )-fuzzy semiprime.
Proof. (i) =) (v) : Let f ,g; h are (2; 2 _qk )-fuzzy generalized bi, fuzzy
generalized interior and fuzzy left ideals of an intra-regular AG-groupoid S
respectively. Now, as S is an intra-regular AG-groupoid so for a 2 S there
exists x; y 2 S such that using left invertive law we get,
a = xa2 y = (x (aa))y = (a(xa))y = a x( xa2 y ) y
= a xa2 (xy) y = y xa2 (xy) a = (y ((x(aa)) (xy))) a
= (y ((a(xa)) (xy))) a = ((a (xa)) (y (xy))) a = (((y (xy)) (xa))a)a
= (aa) ((y(xy)) (xa)) :
Therefore,
((f k g) k h) (a)
_ 1 k
= (f k g) (p) ^ h (q) ^
a=pq
2
( ! )
_ _ 1 k 1 k
= (f (u) ^ g (v)) ^ ^ h (q) ^
a=pq p=uv
2 2
_ 1 k
= (f (u) ^ g (v)) ^ h (q) ^
2
a=(uv)q
_ 1 k
= (f (u) ^ g (v)) ^ h (q) ^
2
a=(uv)q=(aa)((y(xy))(xa))
1k
(f (a) ^ g (a)) ^ h((y(xy)) (xa)) ^
2
1 k 1 k
= (f (a) ^ g (a)) ^ h(a) ^ ^
2 2
1 k
(f (a) ^ g (a)) ^ h(a) ^
2
= ((f ^k g) ^k h) (a) :
Thus, (f ^k g) ^k h (f k g) k h. As given that S is intra-regular so
for a 2 S there exists x; y 2 S such that a = xa2 y: This implies that
g (a) = g xa2 y g a2 :
(v) =) (iv) =) (iii) : are obvious.
(iii) =) (ii) : Assume that B be any bi, R be right and L be left ideal of
S respectively. Now (CB )k , (CR )k and (CL )k are (2; 2 _qk )-fuzzy bi, right
and left ideals of S respectively. Now, by (iii) (CR )k is fuzzy semiprime.
Therefore R is semiprime. Thus we have
(C(B\R)\L )k = (CB ^k CR ) ^k CL (CB k CR ) k CL = (C(BR)L )k :
Hence, (B \ R) \ L (BR)L:
(ii) =) (i) : We know that Sa is both bi and left ideal and Sa2 right
ideal of S containing a and a2 , respectively. And by (ii) Sa2 is semiprime.
So„a 2 Sa2 : Thus, using (ii) ; medial law we have,
a 2 Sa \ Sa2 \ Sa ( Sa)(Sa2 )Sa = ( Sa)S)((Sa2 )S
= ( SS)S)((Sa2 )S = SS)((Sa2 )S = S((Sa2 )S
= (Sa2 )(SS) = (Sa2 )S:
7.4 Regular AG-groupoids

In this section we have characterized regular Abel-Grassmann’s groupoid
in terms of its (2; 2 _qk )-fuzzy ideals.
De…nition 289 An element a of an AG-groupoid S is called regular if
there exist x in S such that a = (ax)a and S is called regular, if every
element of S is regular.
Lemma 290 Let S be an AG-groupoid. If a = a(ax), for some x in S.
Then a = a2 y, for some y in S.
Proof. Using medial law, we get
a = a(ax) = [a(ax)](ax) = (aa)((ax)x) = a2 y, where y = (ax)x.
Lemma 291 Let S be an AG-groupoid with left identity. If a = a2 x, for

some x in S. Then a = (ay)a, for some y in S.
Proof. Using medial law, left invertive law, paramedial law and medial
law, we get
a = a2 x = (aa)x = ((a2 x)(a2 x))x = ((a2 a2 )(xx))x = (xx2 )(a2 a2 )
= a2 ((xx2 )a2 )) = ((xx2 )a2 )a)a = ((aa2 )(xx2 ))a = ((x2 x)(a2 a))a
= [a2 f(x2 x)ag]a = [fa(x2 x)g(aa)]a = [a(fa(x2 x)ga)]a
= (ay)a, where y = fa(x2 x)ga.
Lemma 292 In AG-groupoid S, with left identity, the following holds.
(i) (aS) a2 = (aS) a:

(ii) (aS) ((aS) a) = (aS) a:
(iii) S ((aS) a) = (aS) a:
(iv) (Sa) (aS) = a (aS) :
(v) (aS) (Sa) = (aS) a:
(vi) [a(aS)]S = (aS)a:
(vi) [(Sa)S](Sa) = (aS)(Sa):
(vii) (Sa)S = (aS):
(viii) S(Sa) = Sa:
(ix) Sa2 = a2 S:
Proof. It is easy.
Example 293 Let us consider an AG-groupoid S = f1; 2; 3g in the follow-

ing multiplication table.
1 2 3
1 1 1 1
2 1 1 3
3 1 2 1
It is easy to check that f1; 2g is the quasi-ideal of S. Clearly S is regular
because 1 = 1 1, 2 = (2 3) 2 and 3 = (3 2) 3. Let us de…ne a fuzzy
subset f on8 S as follows:
< 0:9 for x = 1
f (x) = 0:8 for x = 2
:
0:6 for x = 3
Then clearly f is an (2; 2 _qk )-fuzzy ideal of S.
Theorem 294 For an AG-groupoid S; with left identity; the following are
equivalent.
(i) S is regular.
(ii) For bi-ideal B, ideal I and left ideal L of S; B \ I \ L (BI) L:
(iii) B [a] \ I [a] \ L [a] (B [a] I [a]) L [a] ; for some a in S:
Proof. (i) ) (ii)

Assume that B; I and L are bi-ideal, ideal and left ideal of a regular AG-
groupoid S respectively. Let a 2 B \ I \ L: This implies that a 2 I; a 2 B
and a 2 L: Since S is regular so for a 2 S there exist x 2 S such that
using left invertive law and (1) ; we have, a = (ax) a = (((ax) a) x) a =
((xa) (ax)) a = (a ((xa) x)) a = (B ((SI) S)) L = (BI) L:
Thus B \ I \ L (BI) L:
(ii) ) (iii) is obvious.
(iii) ) (i)
B [a] = a [ a2 [ (aS) a; I [a] = a [ Sa [ aS and L [a] = a [ Sa are

principle bi-ideal, principle ideal and principle left ideal of S generated by
a respectively. Thus by left invertive law and paramedial law we have,
a [ a2 [ (aS) a \ (a [ Sa [ aS) \ (a [ Sa)

a [ a2 [ (aS) a (a [ Sa [ aS)
(a [ Sa)
fS (a [ Sa [ aS)g (a [ Sa)
fSa [ S (Sa) [ S (aS)g (a [ Sa)
= (Sa [ aS) (a [ Sa)
= (Sa) a [ (Sa) (Sa) [ (aS) a [ (aS) (Sa)
= a2 S [ a2 S [ (aS) a [ (aS) a
= a2 S [ (aS) a:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) For (2; 2 _qk )-fuzzy bi-ideal f , (2; 2 _qk )-fuzzy ideal g, and (2; 2
_qk )-fuzzy left ideal h of S, (f ^k g) ^k h (f k g) k h.
(iii) For (2; 2 _qk )-fuzzy generalized bi-ideal f , (2; 2 _qk )-fuzzy ideal g,
and (2; 2 _qk )-fuzzy left ideal h of S, (f ^k g) ^k h (f k g) k h.
Proof. (i) ) (iii)

Assume that f , g and h are (2; 2 _qk )-fuzzy generalized bi-ideal, (2; 2
_qk )-fuzzy ideal and (2; 2 _qk )-fuzzy left ideal of a regular AG-groupoid
S, respectively. Now since S is regular so for a 2 S there exist x 2 S
such that using left invertive law we have, a = (ax) a = (((ax) a) x) a =
((xa) (ax)) a = (a ((xa) x)) a. Thus,

_ 1 k
((f k g) k h)(a) = (f k g)(p) ^ h(q) ^
a=pq
2
( ) !
_ _ 1 k 1 k
= f (u) ^ g(v) ^ ^ h(q) ^
a=pq p=uv
2 2
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(uv)q
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(a((xa)x))a=(uv)q
1 k
ff (a) ^ g ((xa) x)g ^ h(a) ^
2
1 k 1 k
f (a) ^ g(a) ^ ^ h(a) ^
2 2
1 k 1 k
= ff (a) ^ g (a) ^ g ^ h (a) ^
2 2
= ((f ^k g) ^k h) (a) :
Therefore (f ^k g) ^k h (f k g) k h:
(iii) ) (ii) is obvious.
(ii) =) (i)
Assume that B, I and L are bi-ideal, ideal and left ideal of S respectively.
Then (CB )k , (CI )k and (CL )k are (2; 2 _qk )-fuzzy bi-ideal, (2; 2 _qk )-
fuzzy ideal and (2; 2 _qk )-fuzzy left ideal of S respectively. Therefore we
have, (CB\I[L )k = (CB ^k CI ) ^k CL (CB k CI ) k CL = (C(BI)L )k =
(C(BI)L )k : Therefore B \ I \ L (BI) L: Hence S is regular.
equivalent.
(i) S is regular.
(ii) For left ideal L, ideal I and quasi-ideal Q of S; L \ I \ Q (LI) Q.
(ii) L [a] \ I [a] \ Q [a] (L [a] I [a]) Q [a] ; for some a in S:
Proof. (i) ) (ii)
Assume that L, I and Q are left ideal, ideal and quasi-ideal of regular
AG-groupoid S. Let a 2 L\I \Q: This implies that a 2 L; a 2 I and a 2 Q:
Now since S is regular so for a 2 S there exist x 2 S such that using left
invertive law and (1) ; we have, a = (ax) a = (((ax) a) x) a = ((xa) (ax)) a =
(a ((xa) x)) a 2 (L ((SI) S)) Q (LI) Q: Thus L \ I \ Q (LI) Q:
(iii) ) (i)
L [a] = a [ Sa, I [a] = a [ Sa [ aS and Q [a] = a [ (Sa \ aS) are left
ideal, ideal and quasi-ideal of S generated a respectively. Thus by medial
law we have,
(a [ Sa) \ (a [ Sa [ aS) \ (a [ (Sa \ aS))
((a [ Sa) (a [ Sa [ aS))
(a [ (Sa \ aS))
f(a [ Sa) Sg (a [ aS)
= faS [ (Sa) Sg (a [ aS)
= (aS) (a [ aS)
= (aS) a [ (aS) (aS)
= (aS) a [ a2 S:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) For (2; 2 _qk )-fuzzy left ideal f , (2; 2 _qk )-fuzzy ideal g, and (2; 2
_qk )-fuzzy quasi-ideal h of S, (f ^k g) ^k h (f k g) k h.
Proof. (i) ) (ii)
Assume that f , g and h are (2; 2 _qk )-fuzzy left ideal, (2; 2 _qk )-fuzzy
ideal and (2; 2 _qk )-fuzzy quasi-ideal of a regular AG-groupoid S, respec-
tively. Now since S is regular so for a 2 S there exist x 2 S such that using
left invertive law; we have, a = (ax) a = (((ax) a) x) a = ((xa) (ax)) a =
(a ((xa) x)) a. Thus,
((f k g) k h)(a)
_ 1 k
= (f k g)(p) ^ h(q) ^
a=pq
2
( ) !
_ _ 1 k 1 k
= f (u) ^ g(v) ^ ^ h(q) ^
a=pq p=uv
2 2
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(uv)q
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(a((xa)x))a=(uv)q
1 k
ff (a) ^ g ((xa) x)g ^ h(a) ^
2
1 k 1 k
f (a) ^ g(a) ^ ^ h(a) ^
2 2
1 k 1 k
= ff (a) ^ g (a) ^ g ^ h (a) ^
2 2
= ((f ^k g) ^k h) (a) :
(ii) =) (i)
Assume that L, I and Q are left ideal, ideal and quasi-ideal of S re-
spectively. Thus (CL )k , (CI )k and (CQ )k are (2; 2 _qk )-fuzzy left ideal,
(2; 2 _qk )-fuzzy ideal and (2; 2 _qk )-fuzzy quasi-ideal of S respectively.
Therefore we have, (CL\I[Q )k = (CL ^k CI ) ^k CQ (CL k CI ) k CQ =
(C(LI)Q )k = (C(LI)Q )k : Therefore L \ I \ Q (LI) Q: Hence S is regular.
equivalent.
(i) S is regular.
(ii) For bi-ideal B, ideal I and quasi-ideal Q of S; B \ I \ Q (BI) Q.
(iii) B [a] \ I [a] \ Q [a] (B [a] I [a]) Q [a] ; for some a in S.
Proof. (i) ) (ii)
Assume that B, I and Q are bi-ideal, ideal and quasi-ideal of regular AG-
groupoid S. Let a 2 B \ I \ Q: This implies that a 2 B; a 2 I and a 2 Q:
(a ((xa) x)) a 2 (B ((SI) S)) Q (BI) Q: Thus B \ I \ Q (BI) Q:
(iii) ) (i)
Since B [a] = a [ a2 [ (aS) a; I [a] = a [ Sa [ aS and Q [a] = a [
(Sa \ aS) are principle bi-ideal, principle ideal and principle quasi-ideal of
S generated by a respectively. Thus by (ii) and medial law and left invertive
law we have,
a [ a2 [ (aS) a \ (a [ Sa [ aS) \ (a [ (Sa \ aS))
( a [ a2 [ (aS) a (a [ Sa
[aS)) (a [ (Sa \ aS))
(S(a [ Sa [ aS)) (a [ aS)
= (Sa [ S (Sa) [ S (aS)) (a [ aS)
= (Sa [ S (Sa) [ S (aS)) (a [ aS)
= (aS [ Sa) (a [ aS)
= (aS) a [ (aS) (aS) [ (Sa) a [ (Sa) (aS)
= (aS) a [ a2 S [ a (aS) :
Hence S is regular.
equivalent.
(i) S is regular.
(ii) For (2; 2 _qk )-fuzzy bi-ideal f , (2; 2 _qk )-fuzzy ideal g, and (2; 2
_qk )-fuzzy quasi ideal h of S, (f ^k g) ^k h (f k g) k h.
(iii) For (2; 2 _qk )-fuzzy generalized bi-ideal f , (2; 2 _qk )-fuzzy ideal g,
and (2; 2 _qk )-fuzzy quasi ideal h of S, (f ^k g) ^k h (f k g) k h.
Proof. (i) ) (iii)

Assume that f , g and h are (2; 2 _qk )-fuzzy generalized bi-ideal, (2; 2
_qk )-fuzzy ideal and (2; 2 _qk )-fuzzy quasi ideal of a regular AG-groupoid
S, respectively. Now since S is regular so for a 2 S there exist x 2 S
such that using left invertive law; we have, a = (ax) a = (((ax) a) x) a =
((xa) (ax)) a = (a ((xa) x)) a. Thus,
((f k g) k h)(a)
_ 1 k
= (f k g)(p) ^ h(q) ^
a=pq
2
( ) !
_ _ 1 k 1 k
= f (u) ^ g(v) ^ ^ h(q) ^
a=pq p=uv
2 2
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(uv)q
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(a((xa)x))a=(uv)q
1 k
ff (a) ^ g ((xa) x)g ^ h(a) ^
2
1 k 1 k
f (a) ^ g(a) ^ ^ h(a) ^
2 2
1 k 1 k
= ff (a) ^ g (a) ^ g ^ h (a) ^
2 2
= ((f ^k g) ^k h) (a) :
(ii) =) (i)
Assume that B, I and Q be bi-ideal, ideal and quasi-ideal of S re-
spectively. Then (CB )k , (CI )k and (CQ )k are (2; 2 _qk )-fuzzy bi-ideal,
Therefore we have, (CB\I[Q )k = (CB ^k CI ) ^k CQ (CB k CI ) k CQ =
(C(BI)Q )k = (C(BI)Q )k : Therefore B \ I \ Q (BI) Q: Hence S is regular.
equivalent.
(i) S is regular.
(ii) For an ideals I1 ; I2 and I3 of S, I1 \ I2 \ I3 (I1 I2 ) I3 .
(iii) I [a] \ I [a] \ I [a] (I [a] I [a]) I [a] ; for some a in S:
Proof. (i) ) (ii)

Assume that I1 , I2 , and I3 are ideals of a regular AG-groupoid S. Let
a 2 I1 \ I2 \ I3 : This implies that a 2 I1 ; a 2 I2 and a 2 I3 : Now since S is
regular so for a 2 S there exist x 2 S such that using left invertive law and
(1) ; we have, a = (ax) a = (((ax) a) x) a = ((xa) (ax)) a = (a ((xa) x)) a 2
(I1 ((SI2 ) S)) I3 (I1 I2 ) I3 : Thus I1 \ I2 \ I3 (I1 I2 ) I3 :
(iii) ) (i)
Since I [a] = a [ Sa [ aS is a principle ideal of S generated by a. Thus
by (iii), left invertive law, medial law and paramedial law we have,
(a [ Sa [ aS) \ (a [ Sa [ aS) \ (a [ Sa [ aS)

((a [ Sa [ aS) (a [ Sa [ aS)) (a [ Sa [ aS)
f(a [ Sa [ aS) Sg (a [ Sa [ aS)
= faS [ (Sa) S [ (aS) Sg (a [ Sa [ aS)
= faS [ Sag (a [ Sa [ aS)
= (aS) a [ (aS) (Sa) [ (aS) (aS) [ (Sa) a
[ (Sa) (Sa) [ (Sa) (aS)
= (aS) a [ a2 S:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) For quasi-ideals Q1 ; Q2 and ideal I of S; Q1 \ I \ Q2 (Q1 I) Q2 .
(iii) Q [a] \ I [a] \ Q [a] (Q [a] I [a]) Q [a] ; for some a in S.
Proof. (i) ) (ii)

Assume that Q1 and Q are quasi-ideal and I is an ideal of a regular AG-
groupoid S. Let a 2 Q1 \I \Q2 : This implies that a 2 Q1 ; a 2 I and a 2 Q2 :
(a ((xa) x)) a 2 (Q1 ((SI) S)) Q2 (Q1 I) Q2 : Thus Q1 \I \Q2 (Q1 I) Q2 .
(iii) ) (i)
Q [a] = a[(Sa \ aS) and I [a] = a[Sa[aS are principle quasi-ideal and
principle ideal of S generated by a respectively. Thus by (iii), left invertive
law, medial law, we have,
(a [ (Sa \ aS)) \ (a [ Sa [ aS) \ (a [ (Sa \ aS))

((a [ (Sa \ aS)) (a [ Sa [ aS))
(a [ (Sa \ aS))
f(a [ aS) Sg (a \ aS)
= faS [ (aS) Sg (a \ aS)
= (aS [ Sa) (a \ aS)
= f(aS) a [ (aS) (aS) [ (Sa) a [ (Sa) aSa
= (aS) a [ a2 S [ a (aS) :
Hence S is regular.
equivalent.
(i) S is regular.
(ii) For (2; 2 _qk )-fuzzy quasi-ideals f; h, and (2; 2 _qk )-fuzzy ideal g,
of S, (f ^k g) ^k h (f k g) k h.
Proof. (i) ) (ii)
Assume that f , h are (2; 2 _qk )-fuzzy quasi-ideal and g is (2; 2 _qk )-
fuzzy ideal of a regular AG-groupoid S, respectively. Now since S is regular
so for a 2 S there exist x 2 S such that using left invertive law; we
have, a = (ax) a = (((ax) a) x) a = ((xa) (ax)) a = (a ((xa) x)) a. Thus,
((f k g) k h)(a)
_ 1 k
= (f k g)(p) ^ h(q) ^
a=pq
2
( ) !
_ _ 1 k 1 k
= f (u) ^ g(v) ^ ^ h(q) ^
a=pq p=uv
2 2
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(uv)q
_ 1 k
= ff (u) ^ g(v)g ^ h(q) ^
2
a=(a((xa)x))a=(uv)q
1 k
ff (a) ^ g ((xa) x)g ^ h(a) ^
2
1 k 1 k
f (a) ^ g(a) ^ ^ h(a) ^
2 2
1 k 1 k
= ff (a) ^ g (a) ^ g ^ h (a) ^
2 2
= ((f ^k g) ^k h) (a) :
(ii) =) (i)
Assume that Q1 and Q2 are quasi-ideals and I is an ideal of S respec-
tively. Thus (CQ1 )k , (CI )k and (CQ2 )k are (2; 2 _qk )-fuzzy quasi-ideal,
Therefore we have,
(CQ1 \I[Q2 )k = (CQ1 ^k CI ) ^k CQ2 (CQ1 k CL ) k CQ2
= (C(Q1 I)Q2 )k = (C(Q1 I)Q2 )k :
Therefore Q1 \ I \ Q2 (Q1 I) Q2 : Hence S is regular.
equivalent.
(i) S is regular.
(ii) For bi-ideal B; B = (BS) B.
(iii) For generalized bi-ideal B; B = (BS) B.
Proof. (i) ) (iii)
Assume that B is generalized bi-ideal of a regular AG-groupoid S: Clearly
(BS) B B: Let b 2 B: Since S is regular so for b 2 S there exist x 2 S
such that b = (bx) b 2 (BS) B: Thus B = (BS) B:
(ii) ) (i)
Since I [a] = a [ a2 [ (aS) a is a principle bi-ideal of S generated by a
respectively. Thus by (ii), we have,
a [ a2 [ (aS) a
= a [ a2 [ (aS) a S a [ a2 [ (aS) a
= f aS [ a2 S [ ((aS) a) S a [ a2 [ (aS) a
= aS [ a2 S [ a (aS) a [ a2 [ (aS) a
= (aS) a [ (aS) a2 [ (aS) ((aS) a)
[ a2 S a [ a2 S a2 [ a2 S ((aS) a)
[ (a (aS)) a [ (a (aS)) a2 [ (a (aS)) ((aS) a)
= (aS) a [ a2 S [ (aS) a [ a2 S [ a2 S [ a2 S
[ (aS) a [ (aS) a [ (aS) a
= a2 S [ (aS) a:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) For (2; 2 _qk )-fuzzy bi-ideal f , of S, fk = (f k S) k f .
(iii) For (2; 2 _qk )-fuzzy generalized bi-ideal f , of S, fk = (f k S) k f .
Proof. (i) ) (iii)

Assume that f is (2; 2 _qk )-fuzzy generalized bi-ideal of a regular AG-
groupoid S. Since S is regular so for b 2 S there exist x 2 S such that
b = (bx) b. Therefore we have,
((f k S) k f )(b)
_ 1 k
= (f k S)(p) ^ f (q) ^
2
b=pq
( ) !
_ _ 1 k 1 k
= f (u) ^ S(v) ^ ^ f (q) ^
p=uv
2 2
b=pq
_ 1 k
= ff (u) ^ S(v)g ^ f (q) ^
2
b=(uv)q
_ 1 k
= ff (u) ^ S(v)g ^ f (q) ^
2
b=(bx)b=(uv)q
1 k
ff (b) ^ S (x)g ^ f (b) ^
2
1 k
f (b) ^ 1 ^ f (b) ^
2
1 k
= f (b) ^ = fk (b) :
2
Thus (f k S) k f fk : Since f is (2; 2 _qk )-fuzzy generalized bi-ideal of

a regular AG-groupoid S: So we have,
((f k S) k f )(b)
_ 1 k
= (f k S)(p) ^ f (q) ^
2
b=pq
( ) !
_ _ 1 k 1 k
= f (u) ^ S(v) ^ ^ f (q) ^
p=uv
2 2
b=pq
( ) !
_ _ 1 k
= f (u) ^ 1 ^ f (q) ^
p=uv
2
b=pq
!
_ _ 1 k
= f (u) ^ f (q) ^
p=uv
2
b=pq
( )
_ _ 1 k
= f (u) ^ f (q) ^
p=uv
2
b==pq
_ 1 k
f ((uv)q) ^
2
b==(uv)q
1 k
= f (b) ^ = fk (b) :
2
This implies that (f k S) k f fk : Thus (f k S) k f = fk :
(ii) =) (i)
Assume that B is a bi-ideal of S. Then (CB )k , is an (2; 2 _qk )-fuzzy bi-
ideal of S. Therefore by by (ii) and, we have, (CB )k = (CB k CS ) k CB =
(C(BS)B )k : Therefore B = (BS) B: Hence S is regular.
8
Interval Valued Fuzzy Ideals of
AG-groupoids
In this chapter we discuss interval valued fuzzy ideals of AG-groupoids.
8.1 Basics
De…nition 305 An interval value fuzzy subset fe on AG-groupoid is called
an interval value (2; 2; _qk) fuzzy AG-subgroupiod of S if xet 2 fe and yse 2 fe
this implies that (xy)minft~;~sg 2 _qk fe for all x; y 2 S and t~; s~ 2 D[0; 1]:
De…nition 306 An interval valued fuzzy subset fe on an AG-groupiod is

called an interval (2; 2 _qk ) fuzzy left(respt right) ideal of an AG -groupiod
of S If yt~ 2 fe This implies that (xy)t~ 2 _qk fe(respt xt~ 2 fe implies that
(xy)t~ 2 _qk fe):
De…nition 307 A fuzzy subset fe of an AG-groupiod S is called an interval

valued (2; 2 _qk )-fuzzy semi prime if x2t~ 2 fe implies that xt~ 2 fe for all
x 2 S:
Theorem 308 An interval value fuzzy subset fe of an AG-groupoid S is

an interval valued (2; 2; _qk )-fuzzy AG-sub groupoid if and only if fe(xy)
minffe(x); fe(y); 1 2 k g:
Proof. Let x; y 2 S and t~; s~ 2 D[0; 1].We assume that x; y 2 S such that
fe(xy) < minffe(x); fe(y)g. we choose t~ 2 D[0; 1]such that fe(xy) < t~
minffe(x); fe(y)g this implies that (xy)t~2 _qk fe and minffe(x); fe(y); 1 2 k g
t~ This implies that fe(x) t~ and fe(y) t~ further xt~ 2 fe and yt~ 2 fe
e
but (xy)t _ qk f which is contradiction due to our wrong supposition so
fe(xy) minffe(x); fe(y); 1 2 k g
Conversely, suppose that fe(xy) minffe(x); fe(y); 1 2 k g. xt~ 2 fe and ys~ 2
fe for t~; s~ 2 D[0; 1] then by de…nition we write it as fe(x) t~ and fe(y) t~
so fe(xy) ffe(x); fe(y)g minft~; 1 2 k g: Here arises two cases:
Case(i): If t~ 1 2 k : Then fe(xy) t~ it mean that (xy)t 2 fe.
Case(ii) If t~ > 1 2 k : Then fe(xy) + t~ + 1 2 k > [1; 1] that as (xy)t~ 2 q fe
From both cases we get (xy)t~ 2 _qk fe: Therefore feis an(2; 2; _qk ) fuzzy
AG-groupiod of S.
Lemma 309 If L is a left ideal if and only if (Cl )k is (2; 2 _qk ) fuzzy left
ideal of S:
Proof. (i) Let x 2 L and s 2 S this implies that xs 2 L now by de…nition
we have (CL )k (x) [1; 1] and (CL )(xs) [1; 1] therefore
1 k
(CL )k (xs) minf(CL )k (x); g
2
(ii)If x 2
= L and s 2 S This implies that xs 2 = L. Then by de…nition we
have (CL )k (x) [0; 0] and (CL )k (xs) [0; 0]
1 k
(CL )k (xs) minf(CL )k (x); g:
2
Conversely let x 2 L; y 2 S Now we have to prove that xy 2 L Then by
de…nition we get (CL )k (x) [1; 1] and now we get
1 k 1 k 1 k
(CL )k (xy) f(CL )k (x); g f[1; 1]; g ;
2 2 2
so we have
1 k
(CL )k (xy) :
2
This implies that xy 2 L:
Theorem 310 An interval valued fuzzy subset fe of an AG-groupiod S is
an interval value (2; 2 _qk ) fuzzy left ideal if and if
1 k
fe(xy) minffe(y); g:
2
Proof. Let x; y 2 S and t~; s~ 2 D[0; 1]. Let fe be an (2; 2 _qk ) fuzzy AG-
groupiod of S on contrary we assume that x; y 2 S Such that fe(xy) fe(y)
we choose t~ 2 D[0; 1] such that
1 k
fe(xy) < t~ minffe(y); g:
2
Then we have fe(y) t~ This implies that yt~ 2 fe and fe(xy) < t~. This implies
that (xy)t~ 2 _qk fe but this is contradiction due to our wrong supposition
Hence
1 k
fe(xy) minffe(y); g:
2
e Now by de…nition we
Conversely let x; y 2 S and t~; s~ 2 D[0; 1] and yt~ 2 f:
have
1 k
fe(xy) minft~; g:
2
Here we consider two cases:
(i) if t~ 1 2 k . Then we have fe(xy) t~ This implies that (xy)t~ 2 fe

(ii) If t~ > 12 k . Then we have
fe(xy) + t~ + k > [1; 1]:
This implies that (xy)t~ 2 qk fe. From both cases we have to prove (xy)t~ 2
_qk fe
Theorem 311 An interval valued fuzzy subset fe of an AG-groupiod S is

called an interval valued (2; 2 _qk ) fuzzy left ideal of S If and only if U (
fe; t~) is left ideal of S for all [0; 0] < t~ 1 2 k :
Proof. Assume that fe is an (2; 2 _qk ) fuzzy left ideal of S: Let us consider
y 2 U ( fe; t~) then fe(y) t~. Then we write fe(xy) minf fe(y); 1 2 k g
minft~; 2 g t~ this implies that fe(xy) t~; this implies that xy 2 U ( fe; t~):
1 k
Hence U ( fe; t~) is left ideal of S:

Conversely Let x; y 2 L and t~ 2 D[0; 1]. Assume that fe(xy) < t~ f
e
f (y); 1 2 k g: Then fe(xy) < t~: This implies that fe(xy) + t~+ k < [1; 1] further
implies that (xy)2 _q k U ( fe; t~) and f fe(y); 1 2 k g t~; fe(y) t~ this implies
that y 2 U ( fe; t~) but xy2 _q k U ( fe; t~): This is contradiction due to our
wrong supposition. Thus fe(xy) minf fe(y); 1 2 k g:
8.2 Main Results using Interval-valued Generalized

Fuzzy Ideals
Theorem 312 Let S be an AG-groupiod with left identity then the follow-
ing condition are equivalent.
(i) S is intra regular.
(ii) For every left ideal L and for any subset I; L \ I LI:
(iii) For every interval-valued (2; 2 _qk ) fuzzy left ideal fe and for every
interval-valued (2; 2 _qk ) be any fuzzy subset g~ then fe ^k g~ fe k g~:
Proof. (i) =) (iii) Assume that S is intra regular AG-groupiod and fe and
g~ are interval-valued (2; 2 _qk ) fuzzy left and interval-valued (2; 2 _qk )
be any fuzzy subset of S: Since S is intra regular therefore for any a in S
Then their exist x; y 2 S such that
a = (xa2 )y = (x(aa)y) = (a(xa))y = y(xa)a:
For any a in S; their exist u and v in S Such that a = uv then we have

1 k
(fe k g~)(a) = _a=uv (fe(u) ^ g~(v) ^ )
2
1 k 1 k
ffe(y(xa) ^ g~(a) ^ g^
2 2
1 k 1 k
ffe(xa) ^ g~(a) ^ g^
2 2
1 k
fe(a) ^ g~(a) ^
2
1 k
= (fe ^ g~)(a) ^
2
= (fe ^k g~)(a):
This implies that fe ^k g~ fe k g~:

(iii) =) (ii) Now let us assume that L be any left ideal and I be any
subset of S: Now (CL )k and (CI )k are the interval-valued (2; 2 _qk ) fuzzy
left and interval-valued (2; 2 _qk ) be fuzzy subset of S: Therefore
(CL\I ) = (CL ^k CI ) CL k CI = (CLI )k (CLI )k :
this implies that L \ I LI:

Now (ii) =) (i) L \ I LI:
a 2 Sa \ Sa (Sa)(Sa) = (SS)(aa) = Sa2 = (Sa2 )S. Hence S is
intra-regular.
Theorem 313 Let S be an AG-groupiod with left identity then following

condition are equivalent.
(ii) For any subset I and for any left ideal L Then I \ L IL:
(iii) For every interval valued (2; 2 _qk ) fuzzy subset fe and for every
interval valued (2; 2 _qk ) fuzzy left ideal g~ then fe ^k g~ fe k g~:
Proof. (i) =) (iii) Let us assume that S is intra regular AG-groupiod

and fe are interval valued (2; 2 _qk ) fuzzy subset and g~ are interval valued
(2; 2 _qk ) fuzzy left ideal of S. Since S is intra regular then for any a 2 S
then their exist x; y 2 S such that
a = (xa2 )y = (x(aa)y = ((a(xa))y = y((xa))a = y(xa) = y(x(xa2 )y)

= y((xa2 ))(xy) = (xa2 )(y(xy) = (xa2 )(xy 2 ) = (y 2 x)(a2 x)
= (y 2 a2 )(x2 ) = (x2 a2 )(y 2 ) = a2 ((x2 y 2 )a):
For any a in S their exist u and v in S Such that a = uv then

1 k
(fe k g~)(a) = _a=uv (fe(u) ^ g~(v)) ^
2
k 1
1 k
ffe(a) ^ g~((x2 y 2 )a) ^
g^
2 2
1 k 1 k
ffe(a) ^ g~(a) ^ g^
2 2
e 1 k
= (f ^ g~)(a) ^
2
= fe ^k g~(a)
fe ^k g~ fe k g~:
(iii) =) (ii) Now let us assume that I be any subset of S and L be any
left ideal of S: Now (CI )k and (CL )k are the (2; 2 _qk ) fuzzy subset and
(2; 2 _qk ) fuzzy left ideal of S: Therefore
(CI\L ) = (CI ^k CL ) CI k CL = (CIL )k :
This implies that I \ L IL:

(ii) =) (i)
a 2 Sa \ Sa (Sa)(Sa) (aa)(SS) = Sa2 = (Sa2 )S.
Hence S is intra regular .
Theorem 314 A fuzzy subset fe of an AG-groupoid S is an interval valued

(2; 2 _qk )-fuzzy semi prime if and only if fe(x) minf fe(x2 ); 1 2 k g;for all
x 2 S:
Proof. Assume that fe is an interval valued (2; 2 _qk )-fuzzy semi prime
so let xe2t 2 fe. This implies that fe(x2 ) t~ Therefore we have fe(x) f
e 2 1 k ~ e ~
f (x ); 2 g = t so f (x) t This implies that xt~ 2 f : e
Conversely let us assume that fe(x) < minf fe(x2 ); 1 2 k g; for all x 2 S:
Then we choose t~ 2 (0; 1]. Now let us assume that fe(x) < t~ minf
fe(x); 1 2 k g then we have fe(x) < t~: This implies that fe(x) + t~ + k < [1; 1]:
Further implies that xet 2 _qk fe and then minf fe(x2 ); 1 2 k g t~ here we
consider fe(x2 ) t~. This implies that x2 2 fe or x2 2 _qk fe. But this implies
e
t e
t
that xet 2 _qk fe: This contradiction arises due to our wrong supposition thus
we have …nal result fe(x) minf fe(x2 ); 1 2 k g:
Example 315 Let S = f1; 2; 3g, and the binary operation “ ”be de…ne on
S as follows:
1 2 3
1 1 1 1
2 1 1 1
3 1 2 1
Then (S, ) is an AG-groupoid. De…ne a fuzzy subset f : S ! [0; 1] as

follows.
8
< 0:77 if x = 1
f (x) = 0:66 if x = 2
:
0:55 if x = 3
Then clearly f is (2; 2 _qk )-fuzzy left ideal.
Example 316 Let S = f1; 2; 3g and binary operation “ ” be de…ned on S

as fallows:
1 2 3
1 2 2 2
2 2 2 2
3 1 2 2
Then (S; ) is an AG-groupoid. De…ne a fuzzy subset f : S ! [0; 1] as

fallows
Theorem 317 Let S be an AG -groupiod with left identity then the fol-
(ii) For any left ideal L and for any subset A of S so A \ L (AL)A:
(iii) For every interval valued (2; 2 _qk )-fuzzy subset fe and every in-
terval valued (2; 2 _qk ) -fuzzy left ideal g of S then fe ^k ge ( fe k ge) k
fe:
Proof. (i) =) (iii) Let fe be the interval valued (2; 2 _qk )-fuzzy subset
and ge be the interval valued (2; 2 _qk )-fuzzy left ideal of an intra regular
AG-groupoid S with left identity then for any a in S their exist x,y 2 S
such that so we use medial law and paramedial law and (ab)c = b(ac)
a = (xa2 )y = (x(aa)y) = (a(xa)y) = (y(xa))a

y(xa) = y(x(xa2 )y) = y((xa2 )(xy)) = (xa2 )(y(xy))
= (xa2 )(xy 2 ) = (xx)(a2 y 2 ) = (x2 )(a2 y 2 ) = a2 (x2 y 2 )
= ((y 2 x2 )a2 ) = ((y 2 x2 )aa)a = a((y 2 x2 )a)a:
For any a in S their exist u and v in S such that a = uv
( fe k ge) k fe(a)
_ 1 k 1 k
= (( fe ge)(u)) ^ fe(v) ^ )^
a=uv
2 2
1 k 1 k
(( fe ge)(a(y 2 x2 )a) ^ fe(a) ^ )^
2 2
1 k 1 k
= (_a((y2 x2 )a)=pq ( fe(p) ^ ge(q)) ^ fe(a) ^ )^
2 2
1 k 1 k
(( fe(a) ^ ge(y 2 x2 )a) ^ fe(a) ^ )^
2 2
1 k 1 k
( fe(a) ^ ge(a)) ^ fe(a) ^ )^
2 2
= ( fe ^k ge)(a) so we have
( fe ^k ge) ( fe k ge) k fe:
(iii) =) (ii) Let A be any subset and L be the left ideal of S then we get
(CA )k are interval valued (2; 2 _qk )-fuzzy subset and (CL )k are interval
valued (2; 2 _qk )-fuzzy left ideal of S then we get
(C(A\L)\A )k = (CA\L )k \ (CA )k (CA k CL ) k CA

= C(AL) k C(A) = C(AL)A :
Hence(A \ L) (AL)A:
(ii) =) (i) Since a is any subset and Sa be the left ideal containing a
so we get the result
a 2 Sa \ Sa (Sa)(Sa) = Sa2 = (Sa2 )S:
Hence S is intra regular.
Theorem 318 Let S be an AG-groupiod with left identity then the follow-
(ii) A \ B AB, for every two sided A and for every bi-ideal B of S:
(iii) For every interval valued (2; 2 _qk )-fuzzy two sided fe and for every
interval value (2; 2 _qk ) -fuzzy bi-ideal ge then fe ^k ge fe k ge:
Proof. (i) =) (iii) Let us assume that S is intra regular AG-groupoid and
fe be interval valued (2; 2 _qk )-fuzzy two sided and ge be interval valued (2
; 2 _qk )-fuzzy bi-ideal of S. Since S is intra regular AG-groupoid therefore
for any a in S. Then their exist x; y 2 S such that
a = (xa2 )y = (x(aa)y) = (a(xa))y = (y(xa))a

y(xa)a = y(x((xa2 )y))a = y((xa2 )(xy) = (xa2 )(y(xy) = a2 (x2 y 2 )
= ((x2 y 2 )a)a = (y 2 x2 )(aa) = a((y 2 x2 )a) = a(ta)
= a(t(xa2 )y)) = a((xa2 )(ty)) = a((yt)(a2 x))
= a(a2 ((yt)x))) = a((yt)x)a)a) = a((ia)a)
= a((i(xa2 )y))a = a((xa2 )(iy))a) = a((yi)(a2 x))a)
= a(a2 ((yi)x))a) = a((x(yi))(aa))a)
= a(a((x(yi)i)a)a) = a((at))a);
so for any a in S their exist u and v in S so a = uv then we get

_ 1 k
( fe k ge)(a) = ( fe(u) ^ ge(v)) ^
a=uv
2
1 k
( fe(a(at )) ^ ge(a)) ^
2
1 k
( fe(a) ^ ge(a)) ^
2
( fe ^k ge)(a)
( fe ^k ge) ( fe k ge):
(iii) =) (ii) Let A be any two sided and B be any bi-ideal of S so we

get (CA )k is interval valued (2; 2 _qk )-fuzzy two sided ideal and (CB )k is
interval valued (2; 2 _qk )-fuzzy bi-ideal of S then we get
(CA\B )k = CA k CB CA k CB (CAB )k :
Hence A \ B AB:
(ii) =) (i) Sa is bi-ideal of an AG-groupoid S containing a and
fag [ fa2 g [ (aS)a is an ideal of S then we get
a 2 (Sa) \ (a [ a2 [ (aS)a) (Sa)(a [ a2 [ (aS)a)

= (Sa)a [ (Sa)a2 [ (Sa)(aS)a Sa2 :

Theorem 319 Let S be an AG-groupoid with left identity then the follow-
ing condition are equivalent
(ii) (Q1 \ Q2 ) \ L (Q1 Q2 )L,for all quasi ideal Q1 and Q2 and left
ideal L of S:
(iii) (fe ^k ge) ^k e
h (fe k ge) k e
h; for all interval valued (2; 2 _qk )-fuzzy
quasi ideals fe and ge and left ideal e
h of S:
Proof. (i) =) (iii) Let us assume that S is intra regular AG-groupoid

with left identity fe and ge are the interval valued (2; 2 _qk ) fuzzy quasi
ideals and e
h be the interval valued (2; 2 _qk ) fuzzy left ideal of S: For each
a in S then their exist x; y 2 S such that
a = (xa2 )y = (x(aa)y) = (a(xa)y) = (y(xa))a = (y(x(xa2 )y))a
= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((xa2 )(xy 2 ))a
= ((y 2 x)(a2 x))a = (a2 (y 2 x)x))a = ((aa)(y 2 x)x))a
= ((a(y 2 x))(ax))a:
Now for any a in S their exist u and v in S such that a = uv then
((fe k ge) k eh)(a)
_ 1 k
= (fe k ge)(u) ^ e h(v) ^
a=uv
2
_ _ 1 k 1 k
= ( fe(p) ^ ge(p) ^ )ê
h(v) ^
a=uv u=pq
2 2
_ 1 k 1 k
= (fe(p) ^ ge(q) ^ )ê
h(v) ^
2 2
a=(pq)v
_ 1 k 1 k
= (fe(p) ^ ge(q) ^ )ê
h(v) ^
2 2
((a(y 2 x))(ax)a
1 k 1 k
(fe(a) ^ fe(a) ^ ge(a) ^ )ê
h(a) ^
2 2
1 k 1 k
((fe(a) ^ ) ^ ge(a)) ^ e
h(a) ^
2 2
1 k 1 k
= (fe(a) ^ ge(a) ^ )ê
h(a) ^
2 2
e e 1 k
= (f ^k ge)(a) ^ h(a) ^
2
e e 1 k
= ((f ^k ge)(a) ^k h(a) ^
2
1 k
= ((fe ^k ge)(a) ^k e
h(a) ^ :
2
Hence (fe ^k ge) ^k e
h (fe k ge) k eh
(iii) =) (ii) Let Q1 ; Q2 and L are the fuzzy quasi ideals and fuzzy left
ideal of S: Then CQ1 and CQ2 and CL are interval valued (2; 2 _qk ) fuzzy
quasi ideals and interval valued fuzzy left ideal of S
(C(Q1 Q2 )L )k (a) =
(CQ1 )k (a) (CQ2 )k (a) (CL )k (a)
((CQ1 ^k CQ2 ) ^k CL )(a)
= (C(Q1 \Q2 )\L)k (a)
Hence (Q1 \ Q2 ) \ L (Q1 Q2 )L:
(ii) =) (i) Let Q and L are the quasi and left ideal of S. Now
a 2 (Sa \ Sa) \ Sa [(Sa)(Sa)](Sa) Sa2 = (Sa2 )S.
(ii) (L1 \ L2 ) \ Q (L1 L2 )Q,L are the fuzzy left ideal and Q are the
fuzzy quasi ideal of S:
(iii) (fe ^k ge) ^k e
h (fe k ge) k e
h; for all interval valued (2; 2 _qk )-fuzzy
left ideals fe and ge and quasi ideal e
h of S:
Proof. (i) =) (iii) Let us assume that S is intra regular AG-groupoid

with left identity fe and ge are the interval valued (2; 2 _qk )-fuzzy left ideals
and eh be the interval valued (2; 2 _qk )-fuzzy quasi ideal of S For each a
in S then their exist x; y 2 S such that
a = (xa2 )y = (x(aa)y) = (a(xa)y) = (y(xa))a = (y(x(xa2 )y))a

= (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a = ((xa2 )(xy 2 ))a
= ((y 2 x)(a2 x))a = (a2 ((y 2 x)x))a = ((aa)(y 2 x))a
= ((a(y 2 x))(ax))a:
Now for any a in S their exist u and v in S such that a = uv then

((fe k ge) k
e
h)(a)
_ 1 k
= (fe k ge)(u) ^ e
h(v) ^
a=uv
2
_ _ 1 k 1 k
= ( fe(p) ^ ge(q) ^ )ê
h(v) ^
a=uv u=pq
2 2
_ 1 k 1 k
= (fe(p) ^ ge(q) ^ )ê
h(v) ^
2 2
a=(pq)v
_ 1 k 1 k
= (fe(p) ^ ge(q) ^ )ê
h(v) ^
2 2
((a(y 2 x))(ax)a
1 k 1 k
(fe(a) ^ fe(a) ^ ge(a) ^ )ê
h(a) ^
2 2
1 k 1 k
((fe(a) ^ ) ^ ge(a)) ^ e
h(a) ^
2 2
1 k 1 k
= (fe(a) ^ fe(a) ^ )ê
h(a) ^
2 2
1 k 1 k
= (fe(a) ^ ge(a) ^ )ê
h(a) ^
2 2
1 k
= ((fe ^k ge)(a) ^k e
h(a) ^
2
1 k
= ((fe ^k ge)(a) ^k e
h(a) ^
2
1 k
= ((fe ^k ge)(a) ^k e
h(a) ^ :
2
Hence (fe ^k ge) ^k e

h (fe k ge) k eh:
(iii) =) (ii) Let L1 ; L2 and Q are the fuzzy left ideals and fuzzy quasi
ideal of S. Then CL1 and CL2 and CQ are interval valued (2; 2 _qk ) fuzzy
left ideals and interval valued fuzzy quasi ideal of S
(C(L1 L2 )Q )k (a) = (CL1 )k (a) (CL2 )k (a) (CQ )k (a)

((CL1 ^k CL2 ) ^k CQ )(a) = (C(L1 \L2 )\Q)k (a):
Hence = (L1 \ L2 ) \ Q (L1 L2 )Q:

(ii) =) (i) Let L and Q are the left and quasi ideal of S. Now
a 2 (Sa \ Sa) \ Sa [(Sa)(Sa)](Sa) = (Sa2 )S:

9
Generalized Fuzzy Ideals of
In this chapter we characterize a Abel-Grassmann’s groupoid in terms of
its (2 ; 2 _q )-fuzzy ideals.
9.1 (2 ; 2 _q )-fuzzy Ideals of AG-groupoids

For the following de…nitions see [65].
Let ; 2 [0; 1] be such that < . For any B A; we de…ne X B be the
fuzzy subset of X by X B (x) for all x 2 B and X B (x) otherwise.
Clearly, X B is the characteristic function of B if = 0 and = 1:
For a fuzzy point xr and a fuzzy subset f of X, we say that
(1) xr 2 f if f (x) r > :
(2) xr q f if f (x) + r > 2 :
(3) xr 2 _q f if xr 2 f or xr q f:
Now we introduce a new relation on F(X), denoted as “ _q( ; ) ”, as
follows.
For any f; g 2 F(X); by f _q( ; ) g we mean that xr 2 f implies
xr 2 _q g for all x 2 X and r 2 ( ; 1]: Moreover f and g are said to be
( ; )-equal, denoted by f =( ; ) g; if f _q( ; ) g and g _q( ; ) f .
Lemma 321 Let f and g are fuzzy subsets of F(X). Then f _q( ; )g if
and only if maxff (x); g minfg(x); g for all x 2 X:
Proof. It is same as in [65]:
Lemma 322 Let f , g and h 2 F(X): If f _q( ; )g and g _q( ; ) h;
then f _q( ; ) h:
Proof. It is same as in [65]:
It is shown in [65] that “=( ; ) ” is equivalence relation on F(X). It
is also noti…ed that f =( ; ) g if and only if maxfminff (x); g; g =
maxfminfg(x); g; g for all x 2 X.
Lemma 323 For an AG-groupoid S, the following holds.
(i) A non empty subset I of AG-groupoid S is an ideal if and only if X I
is ((2 ; 2 _q )-fuzzy ideal.
(ii) A non empty subset L of AG-groupoid S is left ideal if and only if
X L is (2 ; 2 _q )-fuzzy left ideal.
(iii) A non empty subset R of AG-groupoid S is right ideal if and only

if X R is (2 ; 2 _q )-fuzzy right ideal.
(iv) A non empty subset B of AG-groupoid S is bi-ideal if and only if
X B is (2 ; 2 _q )-fuzzy bi-ideal.
(v) A non empty subset Q of AG-groupoid S is quasi-ideal if and only if
X Q is (2 ; 2 _q )-fuzzy quasi-ideal.
Lemma 324 Let A, B be any non empty subsets of an AG -groupoid S
with left identity. Then we have
(1) A B if and only if X A _q( ; ) X B ; where r 2 ( ; 1] and
; 2 [0; 1]:
(2) X A \ X B =( ; ) X (A\B) :
(3) X A X B =( ; ) X (AB) :
Proof. It is same in [65].

Lemma 325 If S is an AG-groupoid with left identity then (ab)2 = a2 b2 =
b2 a2 for all a and b in S.
Proof. It follows by medial and paramedial laws.
; 2 _q )-fuzzy AG-subgroupoid of S if for all x; y 2 S and t; s 2 ( ; 1], it
satis…es xt 2 f; ys 2 f implies that (xy)minft;sg 2 _q f:
Theorem 327 Let f be a fuzzy subset of an AG groupoid S with left iden-
tity. Then f is an (2 ; 2 _q )-fuzzy AG subgroupoid of S if and only
if
maxff (xy); g minff (x); f (y); g where ; 2 [0; 1].
Proof. Let f be a fuzzy subset of an AG-groupoid S which is (2 ; 2 _q )-
fuzzy subgroupoid of S. Assume that there exists x; y 2 S and t 2 ( ; 1],
such that
maxff (xy); g < t minff (x); f (y); g:
Then maxff (xy); g < t. This implies that f (xy) < t, which further im-
plies that (xy)min t 2 _q f and minff (x); f (y); g t. Therefore minff (x); f (y)g
t which implies that f (x) t > , f (y) t > ; implies that xt 2 f ,
ys 2 f: But (xy)minft;sg 2 _q f a contradiction to the de…nition. Hence
maxff (xy); g minff (x); f (y); g for all x; y 2 S:
Conversely, assume that there exist x; y 2 S and t; s 2 ( ; 1] such that

xt 2 f , ys 2 f but (xy)minft;sg 2 _q f; then f (x) t > ; f (y) s > ;
f (xy) < minff (x); f (y); g and f (xy) + minft; sg 2 . It follows that
f (xy) < and so maxff (xy); g < minff (x); f (y); g; this is a contradic-
tion. Hence xt 2 f , ys 2 f implies that (xy)minft;sg 2 _q f for all x; y
in S:
De…nition 328 A fuzzy subset f of an AG-groupoid S with left identity

is called (2 ; 2 _q )-fuzzy left (respt-right) ideal of S if for all x; y 2 S
and t; s ( ; 1] it satis…es yt 2 f implies that (xy)t 2 _q f (respt xt 2 f
implies (xy)t 2 _q f ):
Theorem 329 A fuzzy subset f of an AG-groupoid S with left identity is

called (2 ; 2 _q )-fuzzy left (respt right) ideal of S. if and only if
maxff (xy); g minff (y); g (respt maxff (xy); g minff (x); g).
Proof. Let f be an (2 ; 2 _q )-fuzzy left ideal of S. Let there exists

x; y 2 S and t 2 ( ; 1] such that
maxff (xy); g < t minff (y); g:
Then maxff (xy); g < t this implies that (xy)t 2 f which further
implies that (xy)t 2 _q f . As minff (y); g t > which implies that
f (y) t > , this implies that yt 2 f . But (xy)t 2 _q f a contradiction
to the de…nition. Thus
maxff (xy); g minff (y); g.
Conversely, assume that there exist x; y 2 S and t; s 2 ( ; 1] such that

ys 2 f but (xy)t 2 _q f; then f (y) t > ; f (xy) < minff (y); g
and f (xy) + t 2 . It follows that f (xy) < and so maxff (xy); g <
minff (y); g which is a contradiction. Hence yt 2 f this implies that
(xy)minft;sg 2 _q f (respt xt 2 f implies (xy)minft;sg 2 _q f ) for all
x; y in S:
De…nition 330 A fuzzy subset f of an AG-groupoid S is called (2 ; 2

_q )-fuzzy bi-ideal of S if for all x; y and z 2 S and t; s 2 ( ; 1], the
following conditions hold.
(1) if xt 2 f and ys 2 f implies that (xy)minft;sg 2 _q f:
(2) if xt 2 f and zs 2 f implies that ((xy)z)minft;sg 2 _q f:
Theorem 331 A fuzzy subset f of an AG-groupoid S with left identity is

called (2 ; 2 _q )-fuzzy bi-ideal of S if and only if
(I) maxff (xy); g minff (x); f (y); g:
(II) maxff ((xy)z); g minff (x); f (z); g:
Proof. (1) , (I) is the same as theorem 327.

(2) ) (II) Assume that x; y 2 S and t; s 2 ( ; 1] such that
maxff ((xy)z); g < t minff (x); f (z); g:
Then maxff ((xy)z); g < t which implies that f ((xy)z) < t this im-
plies that ((xy)z)t 2 f which further implies that ((xy)z)t 2 _q f . Also
minff (x); f (z); g t > ; this implies that f (x) t > , f (z) t >
implies that xt 2 f , zt 2 f . But ((xy)z)t 2 _q f; a contradiction. Hence
maxff ((xy)z); g minff (x); f (z); g:
(II) ) (2) Assume that x; y in S and t; s 2 ( ; 1]; such that xt 2

f; zs 2 f but ((xy)z)minft;sg 2 _q f , then f (x) t > ; f (z) s > ;
f ((xy)z) < minff (x); f (y); g and f ((xy)z) + minft; sg 2 . It follows
that f ((xy)z) < and so maxff ((xy)z); g < minff (x); f (y); g a contra-
diction. Hence xt 2 f , zs 2 f implies that ((xy)z)minft;sg 2 _q f for all
x; y in S:
Example 332 Consider an AG-groupoid S = f1; 2; 3g in the following
1 2 3
1 1 1 1
2 1 1 3
3 1 2 1
De…ne a fuzzy
8 subset f on S as follows:
< 0:41 if x = 1;
f (x) = 0:44 if x = 2;
:
0:42 if x = 3:
Then, we have
f is an (20:1 ; 20:1 _q0:11 )-fuzzy left ideal,

f is not an (2; 2 _q0:11 )-fuzzy left ideal,
f is not a fuzzy left ideal.
Example 333 Let S = f1; 2; 3g and the binary operation be de…ned on

S as follows:
1 2 3
1 2 2 2
2 2 2 2
3 1 2 2
Then clearly (S; ) is an AG-groupoid. De…ned a fuzzy subset f on S as
follows: 8
< 0:44 if x = 1;
f (x) = 0:6 if x = 2;
:
0:7 if x = 3:
Then, we have
f is an (20:4 ; 20:4 _q0:45 )-fuzzy left ideal of S.

f is not an (20:4 ; 20:4 _q0:45 )-fuzzy right ideal of S.
equivalent.
(i) S is regular.
(ii) B [a] \ I [a] \ L [a] (B [a] I [a]) L [a] ; for some a in S:
(iii) For bi-ideal B, ideal I and left ideal L of S; B \ I \ L (BI) L:
(iv) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy bi-ideal f , (2
; 2 _q )-fuzzy ideal g, and (2 ; 2 _q )-fuzzy left ideal h of S.
(v) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy generalized
bi-ideal f , (2 ; 2 _q )-fuzzy ideal g, and (2 ; 2 _q )-fuzzy left ideal h of
S:
(vi) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy generalized
bi-ideal f , (2 ; 2 _q )-fuzzy right ideal g, and (2 ; 2 _q )-fuzzy left ideal
h of S:
Proof. (i) ) (vi)

Assume that f , g and h are (2 ; 2 _q )-fuzzy generalized bi-ideal, (2
; 2 _q )-fuzzy right ideal and (2 ; 2 _q )-fuzzy left ideal of a regular AG-
groupoid S, respectively. Now since S is regular so for a 2 S there exist
x 2 S such that using left invertive law and also using law a(bc) = b(ac);
we have,
a = (ax) a = [f(ax) agx]a = (ax)f(ax) ag = [f(ax)agx]f(ax)ag

= f(xa)(ax)gf(ax)ag = [f(ax)ag(ax)](xa)
Thus,
max f((f g) h)(a); g

( )
_
= max f(f g)(x) ^ h(y)g ;
a=xy
maxf(f g)[f(ax)ag(ax)] ^ h(xa); g
_
= maxf (f (u) ^ g(v)) ^ h(xa); g
f(ax)ag(xa)=uv
max ff ((ax)a) ^ g(ax) ^ h(xa); g

= min fmaxff ((ax)a); g ; maxfg(ax); g; maxfh(xa); gg
min fminff (a); g; minfg(a); g ; min{h(a); gg
= min fminff (a) ^ g(a) ^ h(a); g
= min f[f \ g \ h](a); g
Thus f \ g \ h _q( ; ) (f g) h:
(vi) =) (v) is obvious.
(v) =) (iv) is obvious.
(iv) =) (iii)
Assume that B, I and L are bi-ideal, ideal and left ideal of S respectively.
Then B ; I and L are (2 ; 2 _q )-fuzzy bi-ideal, (2 ; 2 _q )-fuzzy
ideal and (2 ; 2 _q )-fuzzy left ideal of S respectively. Therefore we have,
(B\I\L) = ( ; ) B \ I \ L _q( ; )( B I) L
= ( ; )( BI ) L =( ; ) (BI)L :
Therefore B \ I \ L (BI) L:
(ii) ) (i)
B [a] = a [ a2 [ (aS) a; I [a] = a [ Sa [ aS and L [a] = a [ Sa are
principle bi-ideal, principle ideal and principle left ideal of S generated by
a respectively. Thus by (ii), left invertive law, paramedial law and using
law a(bc) = b(ac): we have,
a [ a2 [ (aS) a \ (a [ Sa [ aS) \ (a [ Sa)

a [ a2 [ (aS) a (a [ Sa [ aS) (a [ Sa)
fS (a [ Sa [ aS)g (a [ Sa)
fSa [ S (Sa) [ S (aS)g (a [ Sa)
= (Sa [ aS) (a [ Sa)
= (Sa) a [ (Sa) (Sa) [ (aS) a [ (aS) (Sa)
= a2 S [ a2 S [ (aS) a [ (aS) a
= a2 S [ (aS) a:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) L [a] \ I [a] \ Q [a] (L [a] I [a]) Q [a] ; for some a in S:.
(iii) For left ideal L, ideal I and quasi-ideal Q of S; L \ I \ Q (LI) Q
(iv) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy left ideal f , (2
; 2 _q )-fuzzy ideal g, and (2 ; 2 _q )-fuzzy quasi- ideal h of S.
(v) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy left ideal f , (2
; 2 _q )-fuzzy right ideal g, and (2 ; 2 _q )-fuzzy quasi- ideal h of S.
Proof. (i) ) (ii)

Assume that f , g and h are (2 ; 2 _q )-fuzzy left ideal, (2 ; 2 _q )-
fuzzy right ideal and (2 ; 2 _q )-fuzzy quasi-ideal of a regular AG-groupoid
S, respectively. Now since S is regular so for a 2 S there exist x 2 S such
that using left invertive law and also using law a(bc) = b(ac); we have,
a = (ax) a = [f(ax) agx]a = f(xa)(ax)ga:

Thus

( )
_
= max f(f g)(x) ^ h(y)g ;
a=xy
maxf(f g)f(xa)(ax)g ^ h(a); g
_
= maxf (f (p) ^ g(q)) ^ h(a); g
f(xa)(ax)g=pq
max ff (xa) ^ g(ax) ^ h(a); g

= min fmaxff (xa); g ; maxfg(ax); g; maxfh(a); gg
= min f[f \ g \ h](a); g
Hence f \ g \ h _q( ; ) (f g) h:
(v) ) (iv) is obvious.
(iv) ) (iii)
Assume that L, I and Q are left ideal, ideal and quasi-ideal of S re-
spectively. Then B ; I and L are (2 ; 2 _q )-fuzzy left ideal, (2
; 2 _q )-fuzzy ideal and (2 ; 2 _q )-fuzzy quasi-ideal of S respectively.
Therefore we have,
(L\I\Q) = ( ; ) L \ I \ Q _q( ; )( L I) Q
= ( ; )( LI ) Q =( ; ) (LI)Q :
Therefore L \ I \ Q (LI) Q:
(ii) ) (i)
L [a] = a [ Sa, I [a] = a [ Sa [ aS and Q [a] = a [ (Sa \ aS) are left
ideal, ideal and quasi-ideal of S generated a respectively. Thus by (iii) and
medial law we have,
(a [ Sa) \ (a [ Sa [ aS) \ (a [ (Sa \ aS)) ((a [ Sa) (a [ Sa [ aS))

(a [ (Sa \ aS))
f(a [ Sa) Sg (a [ aS)
= faS [ (Sa) Sg (a [ aS)
= (aS) (a [ aS)
= (aS) a [ (aS) (aS)
= (aS) a [ a2 S:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) B [a] \ I [a] \ Q [a] (B [a] I [a]) Q [a] ; for some a in S.
(iii) For bi-ideal B, ideal I and quasi-ideal Q of S; B \ I \ Q (BI) Q.
(iv) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy bi-ideal f , (2
; 2 _q )-fuzzy ideal g, and (2 ; 2 _q )-fuzzy quasi-ideal h of S:
(v) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy generalized
bi-ideal f , (2 ; 2 _q )-fuzzy ideal g, and (2 ; 2 _q )-fuzzy quasi-ideal h
of S:
Proof. (i) ) (v)
Assume that f , g and h are (2 ; 2 _q )-fuzzy generalized bi-ideal, (2
; 2 _q )-fuzzy ideal and (2 ; 2 _q )-fuzzy quasi-ideal of a regular AG-
groupoid S, respectively. Now since S is regular so for a 2 S there exist
we have,
a = (ax) a = (((ax) a) x) a = ((xa) (ax)) a = [af(xa) xg]a:
Thus,
( )
_
= max f(f g)(b) ^ h(c)g ;
a=bc
maxf(f g)[af(xa) xg] ^ h(a); g
_
= maxf (f (p) ^ g(q)) ^ h(a); g
af(xa)xg=pq
max ff (a) ^ gf(xa)xg ^ h(a); g

= min fmaxff (a); g ; maxfgfxa)xg; g; maxfh(a); gg
= min f[f \ g \ h](a); g
Thus f \ g \ h _q( ; ) (f g) h:
(iv) ) (iii)
Assume that B, I and Q are bi-ideal, ideal and quasi-ideal of regular
AG-groupioud of S respectively. Then B ; I and Q are (2 ; 2 _q )-
fuzzy bi-ideal, (2 ; 2 _q )-fuzzy ideal and (2 ; 2 _q )-fuzzy quasi-ideal
of S respectively. Therefore we have,
(B\I\Q) = ( ; ) L \ I \ Q _q( ; )( B I) Q
= ( ; )( BI ) Q =( ; ) (BI)Q :
Therefore B \ I \ Q (BI) Q:
(ii) ) (i)
Since B [a] = a [ a2 [ (aS) a; I [a] = a [ Sa [ aS and Q [a] = a [
(Sa \ aS) are principle bi-ideal, principle ideal and principle quasi-ideal of
S generated by a respectively. Thus by (ii) and using law a(bc) = b(ac)
medial law and left invertive law we have,
a [ a2 [ (aS) a \ (a [ Sa [ aS) \ (a [ (Sa \ aS))

( a [ a2 [ (aS) a (a [ Sa [ aS)) (a [ (Sa \ aS))
(S(a [ Sa [ aS)) (a [ aS)
= (Sa [ S (Sa) [ S (aS)) (a [ aS)
= (Sa [ S (Sa) [ S (aS)) (a [ aS)
= (aS [ Sa) (a [ aS)
= (aS) a [ (aS) (aS) [ (Sa) a [ (Sa) (aS)
= (aS) a [ a2 S [ a (aS) :
Hence S is regular.
equivalent.
(i) S is regular.
(ii) I [a] \ I [a] \ I [a] (I [a] I [a]) I [a] ; for some a in S:
(iii) For an ideals I1 ; I2 and I3 of S, I1 \ I2 \ I3 (I1 I2 ) I3 .
(iv) f \ g \ h _q( ; ) (f g) h: for any (2 ; 2 _q )-fuzzy ideals f ,g
and h of S.
Proof. (i) ) (iv)

Assume that f , g and h are any (2 ; 2 _q )-fuzzy ideals of a regular
AG-groupoid S, respectively. Now since S is regular so for a 2 S there exist
we have,
a = (ax) a = [f(ax) agx]a = ((xa) (ax)) a:

Thus,

( )
_
= max f(f g)(b) ^ h(c)g ;
a=bc
maxf(f g)f(xa) (ax)g ^ h(a); g
_
= maxf ff (p) ^ g(q)g ^ h(a); g
(xa)(ax)=pq
max ff (xa) ^ g(ax) ^ h(a); g

= min fmaxff (xa); g ; maxfg(ax); g; maxfh(a); gg
= min f[f \ g \ h](a); g
Thus f \ g \ h _q( ; ) (f g) h:
(iv) ) (iii)
Assume that I1 , I2 and I3 are any ideals of regular AG-groupioud of S
respectively. Then I1 ; I2 and I3 are any (2 ; 2 _q )-fuzzy ideals of
S respectively. Therefore we have,
(I1 \I2 \I3 ) = ( ; ) I1 \ I2 \ I3 _q( ; )( I1 I2 ) I3
= ( ; )( I1 I2 ) I3 =( ; ) (I1 I2 )I3 :
Therefore I1 \ I2 \ I3 (I1 I2 ) I3 :
(ii) ) (i)
Since I [a] = a [ Sa [ aS is a principle ideal of S generated by a. Thus
by (ii), left invertive law, medial law and paramedial law we have,
(a [ Sa [ aS) \ (a [ Sa [ aS) \ (a [ Sa [ aS)

((a [ Sa [ aS) (a [ Sa [ aS))
(a [ Sa [ aS)
f(a [ Sa [ aS) Sg (a [ Sa [ aS)
= faS [ (Sa) S [ (aS) Sg (a [ Sa [ aS)
= faS [ Sag (a [ Sa [ aS)
= (aS) a [ (aS) (Sa) [ (aS) (aS) [ (Sa) a
[ (Sa) (Sa) [ (Sa) (aS)
= (aS) a [ a2 S:
Hence S is regular.
equivalent.
(i) S is regular.
(ii) Q [a] \ I [a] \ Q [a] (Q [a] I [a]) Q [a] ; for some a in S.
(iii) For quasi-ideals Q1 ; Q2 and ideal I of S; Q1 \ I \ Q2 (Q1 I) Q2 .
(iv) f \ g \ h _q( ; ) (f g) h: for (2 ; 2 _q )-fuzzy quasi-ideals f
and h, (2 ; 2 _q )-fuzzy ideal g of S.
Proof. (i) ) (iv)

Assume that f , h are (2 ; 2 _q )-fuzzy quasi-ideals and g is (2 ; 2
_q )-fuzzy ideal of a regular AG-groupoid S, respectively. Now since S is
regular so for a 2 S there exist x 2 S such that using left invertive law and
law is a(bc) = b(ac); we have,
a = (ax) a = [f(ax) agx]a = ((xa) (ax)) a = af(xa) xga:
Thus,

( )
_
= max f(f g)(p) ^ h(q)g ;
a=pq
maxf(f g)[af(xa) xg] ^ h(a); g
_
= maxf (f (u) ^ g(v)) ^ h(a); g
af(xa)xg=uv
max ff (a) ^ gf(xa)xg ^ h(a); g

= min fmaxff (a); g ; maxfgfxa)xg; g; maxfh(a); gg
= min f[f \ g \ h](a); g
Thus f \ g \ h _q( ; ) (f g) h:
(iv) ) (iii)
Assume that Q1 and Q2 are quasi-ideals and I is an ideal of a regular
AG-groupoid S. Then Q1 and Q2 are (2 ; 2 _q )-fuzzy quasi-ideal,
I is (2 ; 2 _q )-fuzzy ideal of S. Therefore we have,
(Q1 \I\Q2 ) = ( ; ) Q1 \ I \ Q2 _q( ; )( Q1 I) Q2
= ( ; )( Q1 I ) Q2 =( ; ) (Q1 I)Q2 :
Thus Q1 \ I \ Q2 (Q1 I) Q2 .
(ii) ) (i)
Q [a] = a[(Sa \ aS) and I [a] = a[Sa[aS are principle quasi-ideal and
principle ideal of S generated by a respectively. Thus by (iii), left invertive
law, medial law and we have,
(a [ (Sa \ aS)) \ (a [ Sa [ aS) \ (a [ (Sa \ aS))
((a [ (Sa \ aS)) (a [ Sa [ aS))
(a [ (Sa \ aS))
f(a [ aS) Sg (a \ aS)
= faS [ (aS) Sg (a \ aS)
= (aS [ Sa) (a \ aS)
= f(aS) a [ (aS) (aS) [ (Sa) a [ (Sa) aSa
= (aS) a [ a2 S [ a (aS) :
Hence S is regular.
equivalent.
(i) S is regular.
(ii)For principle bi-ideal B[a]; B[a] = (B[a]S) B[a].
(iii)For bi-ideal B; B = (BS) B.
(iv) For generalized bi-ideal B; B = (BS) B.
(v) For (2 ; 2 _q )-fuzzy bi-ideal f , of S f =( ; ) (f S) f:
(vi) For (2 ; 2 _q )-fuzzy generalized bi-ideal f , of S, f =( ; ) (f S) f:
Proof. (i) ) (vi)
Assume that f is (2 ; 2 _q )-fuzzy generalized bi-ideal of a regular
AG-groupoid S. Since S is regular so for b 2 S there exist x 2 S such that
b = (bx) b. Therefore we have,
max f((f S) f )(b); g
8 9
<_ =
= max f(f S)(x) ^ f (y)g ;
: ;
b=xy
maxf(f s)(bx) ^ f (b); g

_
= maxf (f (u) ^ S(v)) ^ f (b); g
bx=uv
max ff (b) ^ S(x) ^ f (b); g
= min fmaxff (b); g ; 1; maxff (b); gg
minfminff (b); g; 1; min{f (b); gg
= min fminff (b) ^ 1 ^ f (b); g
= min ff (b); g
Thus f _q( ; ) (f S) f: Since f is (2 ; 2 _q )-fuzzy generalized
bi-ideal of a regular AG-groupoid S: So we have (f S) f _q( ; ) f:
Hence f =( ; ) (f S) f:
(vi) ) (v) is obvious.
(v) ) (iv)
Assume that B is a bi-ideal of S. Then B , is an (2 ; 2 _q )-fuzzy
bi-ideal of S. Therefore we have,
B = ( ; )( B S) B
= ( ; )( BS ) B =( ; ) (BS)B :
Therefore B = (BS) B:
(iv) ) (iii) is obvious.
(ii) ) (i)
Since B [a] = a [ a2 [ (aS) a is a principle bi-ideal of S generated by a
respectively. Thus by (ii), we have,
a [ a2 [ (aS) a
= [fa [ a2 [ (aS) agS] a [ a2 [ (aS) a
= [aS [ a2 S [ f(aS) agS] a [ a2 [ (aS) a
= aS [ a2 S [ a (aS) a [ a2 [ (aS) a
= (aS) a [ (aS) a2 [ (aS) ((aS) a)
[ a2 S a [ a2 S a2 [ a2 S ((aS) a)
[ (a (aS)) a [ (a (aS)) a2 [ (a (aS)) ((aS) a)
= (aS) a [ a2 S [ (aS) a [ a2 S [ a2 S [ a2 S
[ (aS) a [ (aS) a [ (aS) a
= a2 S [ (aS) a:
Hence S is regular
equivalent.
(i) S is regular.
(ii) B[a] \ Q [a] (B [a] S) Q [a] ; for some a in S.
(iii) For bi-ideal B and quasi-ideal Q of S; B \ Q (BS) Q.
(iv) f \ g _q( ; ) (f S) g: for (2 ; 2 _q )-fuzzy bi-ideal f , and
(2 ; 2 _q )-fuzzy quasi-ideal g of S:
(v) f \ g _q( ; ) (f S) g: for (2 ; 2 _q )-fuzzy generalized bi-ideal
f and (2 ; 2 _q )-fuzzy quasi-ideal g of S:
Proof. (i) ) (v)
Assume that f and g are (2 ; 2 _q )-fuzzy generalized bi-ideal and (2
; 2 _q )-fuzzy quasi-ideal of a regular AG-groupoid S, respectively. Now
since S is regular so for a 2 S there exist x 2 S such that using left invertive
law and also using law a(bc) = b(ac); we have, a = (ax) a = [f(ax) agx]a:
Thus,
max f((f S) g)(a); g

( )
_
= max f(f S)(b) ^ g(c)g ;
a=bc
maxf(f S)[f(ax)agx] ^ g(a); g
_
= maxf (f (p) ^ S(q)) ^ g(a); g
f(ax)agx=pq
max ff ((ax)a) ^ S(x) ^ g(a); g

= min fmaxff ((ax)a); g ; 1; maxfg(a); gg
minfminff (a); g; 1; min{g(a); gg
= min fminff (a) ^ 1 ^ g(a); g
= min f[f \ g](a); g
Thus f \ g _q( ; ) (f S) g
(iv) ) (iii)
Assume that B and Q are bi-ideal and quasi-ideal of regular AG-groupioud
of S respectively. Then B and Q are (2 ; 2 _q )-fuzzy bi-ideal and
(2 ; 2 _q )-fuzzy quasi-ideal of S respectively. Therefore we have,
(B\Q) = ( ; ) B \ S \ Q _q( ; )( B S) Q
= ( ; )( BS ) Q =( ; ) (BS)Q :
Therefore B \ Q (BS) Q:
(ii) ) (i)
Since B [a] = a [ a2 [ (aS) a and Q [a] = a [ (Sa \ aS) are principle
bi-ideal and principle quasi-ideal of S generated by a respectively. Thus by
(ii) ; law a(bc) = b(ac); medial law and left invertive law we have,
fa [ a2 [ (aS) ag \ fa [ (Sa \ aS)g

f a [ a2 [ (aS) a Sg (a [ (Sa \ aS))
faS [ a2 S [ ((aS) a)Sg (a [ Sa)
= faS [ a2 S [ (Sa)(aS)g (a [ Sa)
= f(aS)a [ (a2 S)a [ f(Sa)(aS)ga [ (aS)(Sa)
[(a2 S)(Sa) [ (Sa)(aS)(Sa)g
(aS)a [ Sa2 [ (aS)a [ (aS)a [ Sa2 [ (aS)a
= (aS) a [ a2 S
Hence S is regular.
9.2 (2 ; 2 _q )-fuzzy Quasi-ideals of

AG-groupoids
The following is an example of generalized fuzzy quasi-ideal in an AG-
groupoid.
Example 341 Consider an AG-groupoid S = f1; 2; 3g in the following

1 2 3
1 3 3 2
2 2 2 2
3 2 2 2
De…ne a fuzzy subset f on S as follows:
8
< 0:21 if x = 1
f (x) = 0:23 if x = 2
:
0:24 if x = 3:
Then, we have
f is an (20:2 ; 20:2 _q0:23 )-fuzzy left ideal,
f is not an (2; 2 _q0:23 )-fuzzy left ideal.

; 2 _q )-fuzzy left (right) ideal of S if it satis…es yt 2 f; (xy)t 2 _q f
(xt 2 f implies that (xy)t 2 _q f ); for all t; s 2 (0; 1]; and ; 2 [0; 1]:
Theorem 343 A fuzzy subset f of an AG-groupoid S is called (2 ; 2

_q )-fuzzy left (respt. right) ideal if and only if maxff (ab); g minff (b); g;
(respt. maxff (ab); g minff (a); g) for all a; b 2 S:
Lemma 344 Every intra regular AG-groupoid S is regular.
Proof. It is easy.
Lemma 345 In an AG-groupoid with left identity S the following holds

(i) (aS)(Sa) = (aS)a, for all a in S,
(ii) f(Sa)(aS)g(Sa) (aS)a, for all a in S.
Proof. (i) Using left invertive law, paramedial law, medial law and 1 we
get
(aS)(Sa) = f(Sa)Sga = f(Sa)(SS)ga = f(SS)(aS)ga = fS(aS)ga = (aS)a.
(ii) now using paramedial and medial laws, and using (i) of this lemma
we get:
f(Sa)(aS)g(Sa) = (aS)f(aS)(Sa)g = (aS)[f(Sa)Sga]

(aS)(Sa) (aS)a.
Lemma 346 In an AG-groupoid with left identity S the following holds

(i) a2 S = (Sa2 )S, for all a in S,
(ii) Sa2 = (Sa2 )S , for all a in S.
Proof. (i) Using (1) we get
a2 S = a2 (SS) = S(a2 S).
(ii)
Sa2 = (SS)a2 = (a2 S)S = f(a2 )(SS)gS = f(SS)(a2 )gS = (Sa2 )S.
Lemma 347 A subset I of an AG-groupoid is left (bi, quasi, two sided)

ideal if and only if X I is (2 ; 2 _q ) fuzzy left(bi, quasi, two sided) ideal.
Proof. It is easy.
Theorem 348 If S is an AG-groupoid with left identity then the following

are equivalent
(i) S is regular,
(ii) B[a] \ L[a] (B[a]S)L[a], for all a in S,
(iii) B \ L (BS)L, where B and L are bi and left ideals of S,
(iv) f \ g _q( ; ) (f X S ) g, where f and g are (2 ; 2 _q )-fuzzy
bi and left ideals of S.
Proof. (i) =) (iv) Let a 2 S, then since S is regular so there exists x in

S such that a = (ax)a. Then using paramedial and medial laws, we get
a = (ax)a = (ax)[(ax)a] = [a(ax)](xa).

maxf(f X S ) g(a); g
" #
_
= max (f X S )(b) ^ g(c) ;
a=bc
max (f X S )(a(ax)) ^ g(xa);
= max min (f X S )(a(ax)); g(xa) ;
2 8 9 3
< _ =
= max 4min f(f (p) ^ X S (q)g; g(xa) ; 5
: ;
a(ax)=pq
max min (f (a); X S (ax); g(xa) ;

= max [min f(f (a); 1; g(xa)g ; ]
= max [min f(f (a); g(xa)g ; ]
= min [maxff (a); g; maxfg(xa); g]
min [minff (a); g; minfg(a); g]
= minf(f \ g)(a); g:
Thus f \ g _q( ; ) (f X S ) g.
(iv) =) (iii) Let B and L are bi and left ideals of S. Then X B and X L
are (2 ; 2 _q )-fuzzy bi and left ideals of S. Now by (iv)
X B\L = X B \X L _q( ; ) (X B X S) X L
= ( ; )
_ q( ; ) (X BS ) X L =( ; )
_q( ; ) X (BS)L .
Thus B \ L (BS)L.
(ii) =) (i) Using left invertive law, paramedial law, medial law, we get
a 2 [a [ a2 [ (aS)a] \ (a [ Sa) [fa [ a2 [ (aS)agS](a [ Sa)

= [aS [ a2 S [ f(aS)agS](a [ Sa)
= (aS)a [ (aS)(Sa) [ (a2 S)a [ (a2 S)(Sa)
[[f(aS)agS]a [ [f(aS)agS](Sa)
(aS)a [ (aS)(Sa) [ Sa2 [ (aS)f(aS)ag
[f(Sa)(aS)g(Sa)
(aS)a [ (Sa2 )S.
Hence S is regular.
are equivalent
(i) S is regular,
(ii) I[a] \ L[a] (I[a]S)L[a] for all a in S,

(iii) I \ L (IS)L for ideal I and left ideal L,
ideal and left ideals of S.

S such that a = (ax)a. Then using paramedial and medial laws, we get
a = (ax)a = (ax)(ea) = (ae)(xa):
Then
" #
_
= max (f X S )(b) ^ g(c) ;
a=bc
" #
_
= max (f X S )(ae) ^ g(xa) ;
a=bc
max (f X S )(ae) ^ g(xa) ;
= max min (f X S )(ae); g(xa) ;
" ( ) #
_
= max min ( f (p) ^ X S (q); g(xa) ;
ae=pq
max min (f (a); X S (e); g(xa) ;

= max [min f(f (a); 1; g(xa)g ; ]
= max [min f(f (a); g(xa)g ; ]
= min [maxff (a); g; maxfg(xa); g]
= minf(f \ g)(a); g:
Thus f \ g _q( ; ) (f X S ) g.
(iv) =) (iii) Let I and L are ideal and left ideal of S respectively. Then
X I and X L are (2 ; 2 _q )-fuzzy ideal and left ideal of S respectively.
Now by (iv)
X I\L = X I \X L _q( ; ) (X I X S) X L
= ( ; )
_ q( ; ) (X IS ) X L =( ; )
_q( ; ) X (IS)L .
Thus I \ L (IS)L.
a 2 (a [ aS [ Sa) \ (a [ Sa) f(a [ aS [ Sa)Sg(a [ Sa)

= faS [ (aS)S [ (Sa)Sg(a [ Sa)
faS [ Sa [ (Sa)Sg(a [ Sa)
= (aS)a [ (Sa)a [ f(Sa)Sga [ (aS)(Sa)
[(Sa)(Sa) [ f(Sa)Sg(Sa)
(aS)a [ (Sa2 )S:
Hence S is regular.

are equivalent
(i) S is regular,
(ii) B[a] (B[a]S)(SB[a]) for all a in S,
(iii) B (BS)(SB), where B is bi-ideal,
(iv) f _q( ; ) (f X S ) (X S f ), where f is fuzzy bi-ideal.
Proof. (i) ) (iv) Let a 2 S, then since S is regular so there exists x in S
such that a = (ax)a. then using medial law we get
a = (ax)a = [f(ax)agx]a = [f(ax)agx](ea)

= [f(ax)age](xa) = [f(ax)age][xf(ax)ag].
max[f(f X S ) (X S f )g(a); ]
" #
_
= max (f X S )(b) ^ (X S f )(c) ;
a=bc
max (f X S )[f(ax)age] ^ (X S f )[xf(ax)ag] ;
= max min (f X S )[f(ax)age]; (X S f )[xf(ax)ag] ;
max min minff f(ax)ag; X S (e)g; minfX S (x); f ((ax)a)g ;
= max [min fminff f(ax)ag; 1g; minf1; f f(ax)aggg ; ]
= max[min ff ((ax)a); f ((ax)a)g ; ]
= min[maxff ((ax)a); g; maxff ((ax)a; )g]
min[minff (a); g; minff ((a); g
= minff (a); g.
Thus f _q( ; ) f(f X S ) (X S f )g.

(iv) =) (iii) Let B be bi-ideal of S. Then X B (2 ; 2 _q )-fuzzy bi-ideal
of S. Now by (iv)
X B _q( ; ) (X B X S )(X S X B) = _q( ; ) X (BS)(SB) .
Thus B (BS)(SB).

(ii) =) (i) Using left invertive law, we get
[a [ a2 [ (aS)a] [fa [ a2 [ (aS)agS][Sfa [ a2 [ (aS)ag]
= [aS [ a2 S [ f(aS)agS][Sa [ Sa2 [ fS((aS)a)g]
= [aS [ a2 S [ (Sa)(aS)][Sa [ Sa2 [ (aS)(Sa)]
(aS)a.
Hence S is regular.
are equivalent
(i) S is regular,
(ii) L[a] \ B[a] (L[a]S)B[a] for all a in S,
(iii) L \ B (LS)B for left ideal L and bi-ideal B,
left and bi-ideals of S respectively.
S such that a = (ax)a.
a = (ax)a = (ax)f(ax)ag.
" #
_
= max (f X S )(b) ^ g(c) ;
a=bc
maxf(f X S) g(a); g
max (f X S )(ax) ^ g((ax)a);
= max min (f X S )(ax); g((ax)a) ;
2 8 9 3
< _ =
= max 4min f(f (p) ^ X S (q)g; g((ax)a) ; 5
: ;
a(ax)=pq
max min minf(f (a); X S (x)g; g((ax)a) ;

= max [min fminf(f (a); 1g; g((ax) a)g ; ]
= max [min f(f (a); g((ax)a)g ; ]
= min [maxff (a); g; maxfg((ax) a); g]
= minf(f \ g)(a); g.
Thus f \ g _q( ; ) (f X S) g.
(iv) =) (iii) Let L and B are ideal and left ideal of S respectively. Then
X L and X B are (2 ; 2 _q )-fuzzy left ideal and bi-ideal of S respectively.
Now by (iv)
X L\B = X L \X B _q( ; ) (X L X S) X B
= ( ; )
_ q( ; ) (X LS ) X B =( ; )
_q( ; ) X (LS)B .
Thus L \ B (LS)B.
a 2 (a [ Sa) \ fa [ a2 [ (aS)ag f(a [ Sa)Sgfa [ a2 [ (aS)ag

= (aS)a [ (aS)a2 [ (aS)f(aS)ag [ f(Sa)Sga
[f(Sa)Sga2 [ f(Sa)Sgf(aS)ag
(aS)a [ (aS)a2 [ (aS)f(aS)Sg [ (aS)(Sa)
[f(Sa)(SS)g(aa) [ [f(aS)agS](Sa)
(aS)a [ (aS)a2 [ (aS)f(aS)Sg [ (aS)(Sa)
[f(Sa)(SS)g(aa) [ f(Sa)(aS)g(Sa)
a(Sa) [ (Sa2 )S.
Hence S is regular.

are equivalent
(i) S is regular,
(ii) L[a] \ Q[a] \ I[a] (L[a]Q[a])I[a] for all a in S,
(iii) L \ Q \ I (LQ)I for left ideal L, quasi-ideal Q and ideal I of S,
(iv) f \ g \ h _q( ; ) (f g) h, where f , g and h are (2 ; 2 _q )-fuzzy
left ideal, right ideal and ideal of S.

S such that a = (ax)a. Now using left invertive law we get
a = (ax)a = [f(ax)agx]a = f(xa)(ax)ga.
maxf(f g) h(a); g
" #
_
= max f(f g)(b) ^ h(a)g ;
a=bc
" #
_
= max f(f g)((xa)(ax)) ^ h(a)g ;
a=bc
max [min f(f g)((xa)(ax)); h(a)g ; ]
2 8 9 3
< _ =
= max 4min f (p) ^ g(q)); h(a) ; 5
: ;
f(xa)(ax)g=pq
max [minfmin ff (xa); g(ax)g; h(a)g ; ]

= max[minff (xa); g(ax); h(a)g; ]
= min [max ff (xa); g; maxfg(ax); g; maxfh(a); g]
min [min ff (a); g; minfg(a); g; minfh(a); g]
= minf(f \ g \ h)(a); g.
Thus f \ g \ h _q( ; ) (f g) h.
(iv) =) (iii) Let L , J and I are left ideal and right ideal and ideal of
S respectively. Then X L , X J and X I are (2 ; 2 _q )-fuzzy left ideal ,
right ideal and ideal of S respectively. Now by (iv)
X L\J\I = X L \X J \X I _q( ; ) (X L X J) X I
= ( ; )
_ q( ; ) (X LJ ) X I =( ; )
_q( ; ) X (LJ)I .
Thus L \ J \ I (LJ)I. Hence L \ Q \ I (LQ)I, where Q is a

quasi-ideal.
a 2 (a [ Sa) \ [a [ f(Sa) \ (aS)g] \ (a [ Sa [ aS)
[(a [ Sa)fa [ f(Sa) \ (aS)gg](a [ Sa [ aS)
= f(a [ Sa)(a [ Sa)g(a [ Sa [ aS)
= fa2 [ a(Sa) [ (Sa)a [ (Sa)(Sa)g(a [ Sa [ aS)
= (a2 )(a) [ (a2 )(Sa) [ (a2 )(aS) [ fa(Sa)ga [ fa(Sa)g(Sa)
[fa(Sa)g(aS) [ f(Sa)aga [ f(Sa)ag(Sa) [ fSa)ag(aS)
[f(Sa)(Sa)ga [ f(Sa)(Sa)g(Sa) [ f(Sa)(Sa)g(aS)
(Sa2 )S [ (Sa)S.
Hence S is regular.

are equivalent
(i) S is regular,
(ii) I[a] \ B[a] I[a](I[a]B[a]) for all a in S,
(iii) I \ B I(IB) for ideal I and bi-ideal B,
(iv) f \ g _q( ; ) f (f g), where f and g are (2 ; 2 _q )-fuzzy ideal

and bi-ideal of S.
S such that a = (ax)a.
a = (ax)a = (ax)f(ax)ag = (ax)[(ax)f(ax)ag].
maxff (f g)(a); g
"( ) #
_
= max f (b) ^ f g(c) ;
a=bc
max [ff (ax) ^ f g(a)g ; ]
= max [min ff (ax); f g(a)g ; ]
" ( ( )) #
_
= max min f (ax); f (p) ^ g(q) ;
a=pq
max [min ff (ax); ff (p) ^ g(q)gg ; ]
= max [min ff (ax); min ff (ax); g((ax)a)gg ; ]
= min [maxff (ax); g; max ff (ax); g; maxfg((ax)a); g]
min [minff (a); g; min ff (a); g; minfg(a); g]
= min ff \ g; g .
Thus f \ g _q( ; ) f (f g).

(iv) =) (iii) Let I and B are ideal and bi-ideal of S respectively. Then
X I and X B are (2 ; 2 _q )-fuzzy ideal and bi-ideal of S respectively.
Now by (iv)
X I\B = X I \X B _q( ; ) (X I ) (X I X B)
= ( ; )
_ q( ; ) (X I ) X IB =( ; )
_q( ; ) X I(IB) .
Thus I \ B I(IB).
(ii) =) (i)Using left invertive law, paramedial law, medial law, we get
(a [ Sa [ aS) \ fa [ a2 [ (aS)ag
(a [ Sa [ aS)[(a [ Sa [ aS)fa [ a2 [ (aS)ag]
= (a [ Sa [ aS)[Sfa [ a2 [ (aS)ag]
= (a [ Sa [ aS)fSa [ Sa2 [ S((aS)a)g
= a(Sa) [ a(Sa2 ) [ a[Sf(aS)ag] [ (Sa)(Sa)
[(Sa)(Sa2 ) [ (Sa)[Sf(aS)ag] [ (aS)(Sa)
[(aS)(Sa2 )(aS)[Sf(aS)ag]
(aS)a [ (Sa2 )S.
Hence S is regular.

are equivalent
(i) S is regular,
(ii) L[a] fL[a](L[a]S)gL[a] for all a in S,
(iii) L fL(LS)gL for left ideal L of S,
(iv) f _q( ; ) ff (f X S )g f where f is (2 ; 2 _q )-fuzzy left ideal
of S.
Proof. (i) =) (iv) Let a 2 S, then since S is regular so there exists x in S

such that a = (ax)a. now using left invertive law, Paramedial law, medial
law and putting xe = x0 ; x0 x = x" we get:
a = (ax)a = [f(ax)agx]a = f(xa)(ax)ga

= fa(ax)g(xa) = f(ea)(ax)g(xa)
= f(xa)(ae)g(xa) = [f(ae)agx](xa)
= x[f((ae)a)xga] = (ex)[f((ae)a)xga]
= [af((ae)a)xg](xe) = [af((ae)a)xg]x0
= [f(ae)ag(ax)]x0 = fx0 (ax)gf(ae)ag
= fa(x0 x)gf(ae)ag = fa(ae)gf(x0 x)ag
= (x0 x)[fa(ae)ga] = (x0 x)[fa(ae)g(ea)]
= (x0 x)[(ae)f(ae)ag] = (ae)[(x0 x)f(ae)ag
= [f(ae)ag(x0 x)](ea) = [f(ae)agx"]a
= [f(x"a)(ae)ga] = [f(ea)(ax")ga]
= [fa(ax")ga].
maxfX S (f f ); g
" #
_
= max ff (f X S )g(b) ^ f (c);
a=bc
max ff (f X S )gfa(ax")g ^ f (a);
= max min ff (f X S )gfa(ax")gg; f (a) ;
2 3
_
= max[min 4[f f (p) ^ (f X S )(q)g; f (a)];
5
fa(ax")gg=pq
max min[minfff (a); (f X S )(ax")g; f (a)g];
= max min[minfff (a); (f X S )(ax")g; f (a)g];

" #
_
= max min[minfff (a); f ff (s) ^ X S (t)gg; f (a)g];
ax"=st
max min[minff (a); ff (a) ^ X S (x")g; f (a)g];
= max min[minff (a); minff (a); X S (x")g; f (a)g];

= max [min[minff (a); minff (a); 1g; f (a)g]; ]
= max [min[minfff (a); f (a); f (a)g]; ]
= max[minff (a); g]
= min[maxff (a); g]
min[minff (a); g].
Thus f _q( ; ) ff (f X S )g f.
(iv) =) (iii) Let L be left ideal of S. Then X L (2 ; 2 _q )-fuzzy left

ideal of S. Now by (iv)
X L _q( ; ) [fX L (X L X S )g X L] = _q( ; ) X [fL(LS)gL] :
Thus L [fL(LS)gL]:
a 2 Sa [(Sa)f(Sa)Sg](Sa)
= [fS(Sa)g(aS)](Sa)
= a[fS(Sa)gS](Sa)
= (aS)(Sa)
(aS)a.
Hence S is regular.

are equivalent
(i) S is regular,
(ii) I[a] \ B[a] I[a](SB[a]) for all a in S,
(iii) I \ B I(SB) for left ideal I and bi-ideal B,
(iv) f \ g _q( ; ) f (X S g), where f and g are (2 ; 2 _q )-fuzzy
interior ideal and bi-ideal of S.
Proof. (i) =) (iv) Let a 2 S, then since S is regular so using left invertive
law we get
a = (ax)a = f(ea)xga = f(xa)egf(ax)ag

= (ax)[f(xa)ega] = (ax)f(ae)(xa)g
= (ax)[xf(ae)ag].
maxff (X S g); g
"( ) #
_
= max f (p) ^ (X S g)(q) ;
a=pq
max f (ax) ^ (X S g)[xf(ae)ag];

= max minff (ax); (X S g)fx((ae)a)gg;
2 3
_
= max 4minff (ax); f (X S (s) ^ g(t); 5
x((ae)a)g=st
max minff (ax); f(X S (x) ^ g((ae)a)g;

= max fminff (ax); minf(X S (x); g((ae)a)gg;
= max [fminff (ax); minf1; g((ae)a)gg; ]
= max [fminff (ax); g((ae)a)g; ]
= min [maxff (ax); g; maxfg((ae)a); g]
min[minff (a); g; minfg(a); g
= minff \ g(a); g.
Thus f \ g _q( ; ) f (X S g).

(iv) =) (iii) Let I and B are ideal and bi-ideal of S respectively. Then
X I and X B are (2 ; 2 _q )-fuzzy ideal and bi-ideal of S respectively.
Now by (iv)
X I\B = X I \X B _q( ; )X I (X S X B)
= ( ; )
_ q( ; ) X fI(SB)g :
Thus I \ B I(SB).
(ii) =) (i)Using fS(Sa)g (Sa) and we get
a 2 (aS [ Sa) \ fa [ a2 [ (aS)ag

(aS [ Sa)[Sfa [ a2 [ (aS)ag]
= (aS [ Sa)[Sa [ Sa2 [ Sf(aS)ag]
= (aS [ Sa)fSa [ Sa2 [ (aS)(Sa)g
= (aS)(Sa) [ (aS)(Sa2 ) [ (aS)f(aS)(Sa)g
[(Sa)(Sa) [ (Sa)(Sa2 ) [ (Sa)f(aS)(Sa)g
(aS)a.
Hence S is regular.
10
On Fuzzy Soft Intra-regular
In this chapter we characterize intra-regular AG-groupoids in terms of (2
; 2 _q )-fuzzy soft ideals.
De…nition 356 Let S be an AG-groupoid and U be an initial universe and

let E be a set of parameters. A pair (F; E) is called a soft set over U if and
only if F is a mapping of E into the set of all subsets of U .
Generally, the soft set, i.e, a pair (F; A) with A B and F : A ! P (S):
De…nition 357 Let (F; A) and (G; B) be soft sets over S, then (G; B) is
called a soft subset of (F; A) if B A and G(b) F (b) for all b 2 B:
Generally we write it as (G; B) ~ (F; A): (F; A) is the soft supperset of

(G; B); if (G; B) is a soft subset of (F; A):
De…nition 358 A soft set (F; A) over an AG-groupoid S is called a soft

AG-groupoid over S if (F; A) (F; A) (F; A):
De…nition 359 A soft set (F; A) over an AG-groupoid S is called a soft

left (right)ideal over S, (S; E) (F; A) (F; A)((F; A) (S; E)
(F; A)):
A soft set over S is a soft ideal if it is both a soft left and a soft right
ideal over S.
De…nition 360 Let V U: A fuzzy soft set hF; Ai over V is said to be

a relative whole ( ; )- fuzzy soft set (with respect to universe set V and
parameter set A), denoted by (V; A); if F (") = f V for all " 2 A:
De…nition 361 A new ordering relation is de…ned on F(X) denoted as

" _q( ; ) ";as follows.
For any ; 2 F(X); by _q( ; ) , we mean that xr 2 implies
xr 2 _q for all x 2 X and r 2 ( ; 1]:
De…nition 362 Let hF; Ai and hG; Bi be two fuzzy soft sets over U. We
say that hF; Ai is an ( ; )-fuzzy soft subset of hG; Bi and writehF; Ai ( ; )
hG; Bi if
(i) A B;
(ii) For any " 2 A; F (") _q( ; ) G("):
De…nition 363 For any fuzzy soft set hF; Aiover an AG-groupiod S, 2 A
and r 2 ( ; 1], denote F ( )r = fx 2 S jxr 2 F ( )g, hF ( )ir = fx 2 S jxr 2 q F ( )g,
[F ( )]r = fx 2 S jxr 2 _q F ( )g :
De…nition 364 Suppose f be a fuzzy subset of an AG-groupoid S; A 2
[0; 1] : De…ne the map F : A ! P (S) as
F ( ) = fx 2 S : f (x) g for all 2 A:
Indeed F ( ) is parameterized family of -level subsets, corresponding
to f: Therefore (F; A) is a soft set over S.
We also de…ne another map, Fq : A ! P (S) as follows
Fq ( ) = fx 2 S : f (x) + > 1g for all 2 A: Then (Fq ; A) is a soft
set over S:
De…ne a map F : A ! P (S) as follows
F ( ) = fx 2 S : f (x) > g for all 2 A: Therefore (F ; A) is a soft
set over S:
Example 365 Let S = fa; b; c; dg and the binary operation " " de…nes on
S as follows:
a b c d
a a a a a
b a d d c
c a d d d
d a d d d
Then (S; ) is an AG-groupoid. Let E = f0:3; 0:4g and de…ne a fuzzy soft
set hG; Ai over S as follows:
2 if x 2 fa; bg
G( )(x) = 1
2 otherwise
Then hG; Ai is an (20:2 ; 20:2 _q0:4 )- fuzzy soft left ideal of S.
Let E = f0:6; 0:7g and de…ne a fuzzy soft set hF; Ai over S as follows:
if x 2 fa; bg
F ( )(x) = 1
2 otherwise
Then hF; Ai is an (20:3 ; 20:3 _q0:4 )- fuzzy soft bi-ideal of S.
Theorem 366 A fuzzy subset f of an AG-groupoid S is fuzzy interior ideal
if and only if (F; A) is a soft interior ideal of S where A = [0; 1] :
Proof. Let f be a fuzzy interior ideal of S then for all x; a; y 2 S; f ((xa) y)
f (a) : Now let a 2 F ( ) this implies that fa 2 S : f (a) g for all 2 A:
This implies that f (a) implies that f ((xa) y) f (a) implies that
((xa) y) 2 F ( ) implies that F ( ) is an interior ideal implies that (F; A)
is soft interior ideal.
Conversely, let (F; A) is soft interior ideal of S we show that f is fuzzy
interior ideal of S. Let f ((xa) y) < f (a) for some x; y; a 2 S and choose
2 A such that f ((xa) y) < f (a) this implies that a 2 F ( ) but

(xa) y 2
= F ( ) which is a contradiction. Hence f is fuzzy interior ideal of
S.
Theorem 367 A fuzzy subset f of a AG-groupoid S is fuzzy bi-ideal if and
only if (F; A) is a soft bi-ideal of S where A = [0; 1] :
Proof. Let f be a fuzzy bi-ideal of S then for all x; y; z 2 S; f ((xy) z)
f (x) ^ f (z) : Now let x; z 2 F ( ) this implies that
fx; y 2 S : f (x) ; f (z) g for all 2 A:
This implies that f (x) ; f (z) implies that f (x) ^ f (z)

implies that f ((xy) z) implies that ((xy) z) 2 F ( ) implies that F ( )
is an bi-ideal implies that (F; A) is soft bi-ideal over S.
Conversely, let (F; A) is soft bi-ideal of S: we show that f is fuzzy bi-
ideal of S. Let f ((xy) z) < f (x) ^ f (z) for some x; y; z 2 S and choose
2 A such that f ((xy) z) < f (x) ^ f (z) this implies that x; z 2 F ( )
but (xy) z 2= F ( ) which is a contradiction. Hence f is fuzzy bi-ideal of S.
Theorem 368 A fuzzy subset f of an AG-groupoid S is fuzzy interior ideal

if and only if (Fq ; A) is a soft interior ideal of S where A = [0; 1] :
Proof. Let f be a fuzzy interior ideal of S then for all x; a; y 2 S; f ((xa) y)
f (a) : Now let a 2 F ( ) this implies that fa 2 S : f (a) + > 1g for all
2 A: This implies that f (a) + > 1 implies that f ((xa) y) + > 1 im-
plies that ((xa) y) 2 Fq ( ) implies that Fq ( ) is an interior ideal implies
that (Fq ; A) is soft interior ideal.
Conversely, let (Fq ; A) is soft interior ideal of S we show that f is fuzzy
interior ideal of S. Let f ((xa) y) < f (a) for some x; a; y 2 S and choose
2 A such that f ((xa) y) < f (a) this implies that a 2 F ( ) but
(xa) y 2= F ( ) which is a contradiction. Hence f is fuzzy interior ideal of
S.
Theorem 369 A fuzzy subset f of an AG-groupoid S is fuzzy bi-ideal if
and only if (Fq ; A) is a soft bi-ideal of S where A = [0; 1] :
Proof. Let f be a fuzzy bi-ideal of S then for all x; y; z 2 S; f ((xy) z)
f (x) ^ f (z) : Now let x; z 2 Fq ( ) this implies that
fx; z 2 S : f (x) + 1; f (z) + 1g for all 2 A:
This implies that f (x)+ 1 , f (z)+ 1 implies that f (x)^f (z)+

1 implies that f ((xy) z) + 1 implies that ((xy) z) 2 Fq ( ) implies that
Fq ( ) is an bi-ideal implies that (Fq ; A) is soft bi-ideal over S.
Conversely, let (Fq ; A) is soft bi-ideal of S we show that f is fuzzy bi-
ideal of S. Let f ((xy) z) < f (x) ^ f (z) for some x; y; z 2 S and choose
2 A such that f ((xy) z) + < 1 f (x) ^ f (z) + this implies that

x; z 2 F ( ) but (xy) z 2
= F ( ) which is a contradiction. Hence f is fuzzy
bi-ideal of S.
Theorem 370 Let f be a fuzzy subset of an AG-groupoid S; then (F; (0:5; 1])
is a soft interior ideal if and only if max ff ((xa) y) ; 0:5g f (a) :
Proof. Let (F; (0:5; 1]) be a soft interior ideal over S, then F ( ) is an
interior ideal of S for each 2 (0:5; 1] such that max ff ((xa) y) ; 0:5g <
f (a) : Choose an 2 (0:5; 1] such that max ff ((xa) y) ; 0:5g < < f (a).
Then a 2 F ( ) but ((xa) y) 2 = F ( ) which is a contradiction,therefore
max ff ((xa) y) ; 0:5g f (a) :
Conversely, let max ff ((xa) y) ; 0:5g f (a) and (F; (0:5; 1]) be a soft
set over S: Let a 2 F ( ) ;where 2 (0:5; 1]: Then max ff ((xa) y) ; 0:5g
f (a) > 0:5: So ((xa) y) 2 F ( ) : Therefore F ( ) is an interior ideal
of S. Hence (F; (0:5; 1]) is a soft interior ideal over S:
Theorem 371 Let f be a fuzzy subset of an AG-groupoid S; then (F; (0:5; 1])
is a soft bi-ideal if and only if max ff ((xy) z) ; 0:5g f (x) ^ f (z) :
Proof. Let (F; (0:5; 1]) be a soft bi-ideal over S, then F ( ) is an bi-ideal
of S for each 2 (0:5; 1] such that max ff ((xy) z) ; 0:5g < f (x) ^ f (z) :
Choose an 2 (0:5; 1] such that max ff ((xy) z) ; 0:5g < < f (x) ^ f (z).
Then x; z 2 F ( ) but ((xy) z) 2 = F ( ) which is a contradiction,therefore
max ff ((xy) z) ; 0:5g f (x) ^ f (z) :
Conversely, let max ff ((xy) z) ; 0:5g f (x) ^ f (z) and (F; (0:5; 1]) be a
soft set over S: Let x; z 2 F ( ) ;where 2 (0:5; 1]: Then max ff ((xy) z) ; 0:5g
f (x) ^ f (z) > 0:5: So ((xy) z) 2 F ( ) : Therefore F ( ) is an bi-ideal
of S. Hence (F; (0:5; 1]) is a soft bi-ideal over S:
interior ideal of S if and only if (F; (0; 0:5]) is a soft interior ideal over S.
Proof. Let f be an (2; 2 _q)-fuzzy interior ideal of S, for x; a; y 2 S;
f ((xa) y) f (a) ^ 0:5:Now let a 2 F ( ) ; then f (a) so a 2 f this
implies that ((xa) y) 2 _qf that is f ((xa) y) or f ((xa) y) + > 1: If
f ((xa) y) then ((xa) y) 2 F ( ) : If f ((xa) y) + > 1 then f ((xa) y) >
1 because 2 (0; 0:5]: So ((xa) y) 2 F ( ) :Thus F ( ) is an interior
ideal of S for all 2 (0; 0:5]: Consequently (F; (0; 0:5]) is a soft interior ideal
over S.
Conversely, Suppose that (F; (0; 0:5]) is a soft interior ideal over S. Then
F ( ) is an interior ideal of S for all 2 (0; 0:5]: We have to show that f
is an (2; 2 _q)-fuzzy interior ideal of S: If possible let there exists some
x; a; y 2 S such that f ((xa) y) < f (a) ^ 0:5: Choose an 2 (0; 0:5] such
that f ((xa) y) < < f (a)^0:5; this shows that a 2 F ( ) but ((xa) y) 2 = F:
Which is a contradiction, Thus f ((xa) y) f (a)^0:5: Thus f is (2; 2 _q)-
fuzzy interior ideal of S:

bi-ideal of S if and only if (F; (0; 0:5]) is a soft bi-ideal over S.
Proof. Let f be an (2; 2 _q)-fuzzy bi-ideal of S, for x; y; z 2 S; f ((xy) z)

f (x) ^ f (z) ^ 0:5:Now let x; z 2 F ( ) ; then f (x) and f (z)
so x ; z 2 f this implies that ((xy) z) 2 _qf that is f ((xy) z) or
f ((xy) z)+ > 1: If f ((xy) z) then (xy) z 2 F ( ) : If f ((xy) z)+ > 1
then f ((xy) z) > 1 because 2 (0; 0:5]: So (xy) z 2 F ( ) : Thus
F ( ) is bi-ideal of S for all 2 (0; 0:5]: Consequently (F; (0; 0:5]) is a soft
bi-ideal over S.
Conversely, assume that (F; (0; 0:5]) is a soft bi-ideal over S. Then F ( )
is an bi-ideal of S for all 2 (0; 0:5]: We have to show that f is an (2
; 2 _q)-fuzzy bi-ideal of S: If possible let there exists some x; y; z 2 S
such that f ((xy) z) < f (x) ^ f (z) ^ 0:5: Choose an 2 (0; 0:5] such
thatf ((xy) z) < < f (x) ^ f (z) ^ 0:5; this shows that x; z 2 F ( ) but
((xy) z) 2
= F: Which is a contradiction, thus f ((xy) z) f (x) ^ f (z) ^ 0:5:
Thus f is (2; 2 _q)-fuzzy bi-ideal of S:
Theorem 374 Let f be a fuzzy subset of an AG-groupoid S. Then f is a

(q; q)-fuzzy interior ideal if and only if (Fq ; (0:5; 1]) is a soft interior ideal
over S.
Proof. Let f be an (q; q)-fuzzy interior ideal of S and suppose that a 2

Fq ( ) where 2 (0:5; 1], then f (a) + 1; that is a qf . Then for each
x; y 2 S; ((xa) y) qf That is f ((xa) y) + 1: Hence ((xa) y) 2 Fq ( ) :
Thus Fq ( ) is an interior ideal of S. Consequently (Fq ; (0:5; 1]) is a soft
interior ideal over S.
Conversely suppose that (Fq ; (0:5; 1]) is a soft interior ideal over S. As-
sume that there exists some a 2 S and 2 (0:5; 1], such that a qf but
((xa) y) q f that is f ((xa) y) + < 1 f (a) + for some x; a; y 2 S:
Then a 2 Fq ( ) but ((xa) y) 2 = Fq ( ), which is a contradiction therefore
f ((xa) y) + 1: Hence ((xa) y) qf: Which shows that f is a (q; q)-fuzzy
interior ideal of S.
Theorem 375 Let f be a fuzzy subset of an AG-groupoid S. Then f is a

(q; q)-fuzzy bi-ideal if and only if (Fq ; (0:5; 1]) is a soft bi-ideal over S.
Proof. Let f be an (q; q)-fuzzy bi-ideal of S and suppose that x; z 2

Fq ( ) where 2 (0:5; 1], then f (x) + 1 and f (z) + 1; that is
x qf;and.z qf . Then for each x; z 2 S; ((xy) z) qf That is f ((xy) z)+
1: Hence ((xy) z) 2 Fq ( ) : Thus Fq ( ) is an bi-ideal of S. Consequently
(Fq ; (0:5; 1]) is a soft bi-ideal over S.
Conversely suppose (Fq ; (0:5; 1]) is a soft bi-ideal over S. Assume that
there exists some x; z 2 S and 2 (0:5; 1], such that x qf and.z qf but
((xy) z) qf that is f ((xy) z)+ < 1 f (x)^f (z)+ for some x; y; z 2 S:
Then x; z 2 Fq ( ) but ((xy) z) 2 = Fq ( ), which is a contradiction therefore
f ((xy) z) + 1: Hence ((xy) z) qf: Which shows that f is a (q; q)-fuzzy

bi-ideal of S.
De…nition 376 The restricted product (H; C) of two soft sets (F; A) and
(G; B) over a semigroup S is de…ned as the soft set (H; C) = (F; A) (G; B)
where C = A \ B and H is a set valued function from C to P (S) de…ned
as H(c) = F (c) G(c) for all c 2 C:
De…nition 377 Let X be a non empty set. A fuzzy subset f of X is de…ned

as a mapping from X into [0; 1], where [0; 1] is the usual interval of real
numbers. The set of all fuzzy subsets of X is denoted by F(X):
De…nition 378 A fuzzy subset f of X of the form

r(6= 0 if y = x;
f (y) =
0 otherwise
is said to be a fuzzy point with support x and value r and is denoted by

xr ; where r 2 (0; 1]:
Let ; 2 [0; 1] be such that < : For any Y X;we de…ne Y be the
fuzzy subset of X by Y (x) for all x 2 Y and Y (x) otherwise.
Clearly, Y is the characteristic function of Y if = 0 and = 1:
For a fuzzy point xr and a fuzzy subset f of X, we say that
(i) xr 2 f if f (x) r > :
(ii) xr q f if f (x) + r > 2 :
(iii) xr 2 _q f if xr 2 or xr q f:
(iiii) xr 2 ^q f if xr 2 and xr q f:
De…nition 379 Let S be an AG-groupoid and ; 2 F(S) .De…ne the

product of and ; denoted by ; by
supx=yz min f (y); (z)g if there exist y; z 2 S such that x = yz;
( )(x) =
0; otherwise.
for all x 2 S:
The following de…nitions are basics are available in [16].
De…nition 380 A pair hF; Ai is called fuzzy soft set over U , where A E
and F is a mapping given by F : A ! F(U ):
In general, for every " 2 A; F (") is a fuzzy set of U and it is called

fuzzy value set of parameter ": The set of all fuzzy soft sets over U with
parameters from E is called a fuzzy soft class, and it is denoted by F (U; E):
De…nition 381 Let hF; Ai and hG; Bi be two soft sets over U . We say
that hF; Ai is a fuzzy soft subset of hG; Bi and write hF; Ai b hG; Bi if
(i) A B;
(ii) For any " 2 A; F (") G("):
hF; Ai and hG; Bi are said to be fuzzy soft equal and write hF; Ai =
hG; Bi if hF; Ai b hG; Bi and hG; Bi b hF; Ai:
De…nition 382 The extended intersection of two fuzzy soft sets hF; Ai and
hG; Bi over U is called fuzzy soft set denoted by hH; Ci; where C = A [ B
and
8
< F (") if " 2 A B;
H(") = G( ") if " 2 B A;
:
F (") \ G(") if " 2 A \ B;
for all " 2 C: This is denoted by hH; Ci = hF; Ai \ hG; Bi:

De…nition 383 The extended union of two fuzzy soft sets hF; Ai and hG; Bi
over U is a fuzzy soft set denoted by hH; Ci; where C = A [ B and
8
< F (") if " 2 A B;
H(") = G( ") if " 2 B A;
:
F (") [ G(") if " 2 A \ B;
for all " 2 C: This is denoted by hH; Ci = hF; Ai [ hG; Bi:

De…nition 384 Let hF; Ai and hG; Bi be two fuzzy soft sets over U such
that A \ B 6= : The restricted intersection of hF; Ai and hG; Bi is de…ned
to be fuzzy soft set hH; Ci; where C = A \ B and H(") = F (") \ G(") for
all " 2 C:This is denoted by hH; Ci = hF; Ai \ hG; Bi:
De…nition 385 Let hF; Ai and hG; Bi be two fuzzy soft sets over U such
that A \ B 6= : The restricted union of hF; Ai and hG; Bi is de…ned to be
fuzzy soft set hH; Ci; where C = A \ B and H(") = F (") [ G(") for all
" 2 C. This is denoted by hH; Ci = hF; Ai [ hG; Bi:
De…nition 386 The product of two fuzzy soft sets hF; Ai and hG; Bi over
an semigroup S is a fuzzy soft set over S, denoted by hF G; Ci; where
C = A [ B and
8
< F (") if " 2 A B;
(F G)(") = G( ") if " 2 B A;
:
F (") G(") if " 2 A \ B;
for all " 2 C: This is denoted by hF G; Ci = hF; Ai hG; Bi:
De…nition 387 A fuzzy soft set hF; Ai over an AG-groupoid is called an
(2 ; 2 _q )-fuzzy soft left (resp., right) ideal over S if it satis…es
(S; A) hF; Ai b( ; ) hF; Ai (resp:; hF; Ai (S; A) b( ; ) hF; Ai)
A fuzzy soft set over S is called (2 ; 2 _q )-fuzzy soft ideal over S if

it is both an (2 ; 2 _q )-fuzzy soft left ideal and an (2 ; 2 _q )-fuzzy
soft right ideal over S.
De…nition 388 A fuzzy soft set hF; Ai over an AG-groupoid S is called

an (2 ; 2 _q )-fuzzy soft bi-ideal over S if it satis…es
(i) hF; Ai hF; Ai b( ; ) hF; Ai;

(ii) hF; Ai (S; A) hF; Ai b( ; ) hF; Ai:
De…nition 389 A fuzzy soft set hF; Ai over an AG-groupoid S is called

(2 ; 2 _q ) fuzzy soft quasi-ideal over S if it satis…es
hF; Ai (S; A) \ (S; A) hF; Ai b( ; ) hF; Ai
Theorem 390 A fuzzy set f over an AG-groupoid S is called (2 ; 2

_q ) fuzzy left (resp. right) ideal over S if it satis…es
(for all x; y 2 S)(max ff (xy); g min ff (y); g (resp: min ff (x); g)):
Proof. It is same as in .
De…nition 391 A fuzzy set f over an AG-groupoid S is called (2 ; 2

_q ) fuzzy left ideal over S if
(f or all x, y 2 S)(f or all t; 2 ( ; 1))(yt 2 f =) (xy)minft;sg 2
_q f ):

_q ) fuzzy right ideal over S if
(f or all x, y 2 S)(f or all t; 2 ( ; 1))(xt 2 f =) (xy)minft;sg 2
_q f ):

_q ) fuzzy ideal over S if it is both (2 ; 2 _q ) fuzzy left ideal and
(2 ; 2 _q ) fuzzy right ideal.

_q ) fuzzy bi-ideal over S if it satis…es
(i) (f or all x; y 2 S)(max ff (xy); g min ff (x); f (y); g);
(ii) (f or all x; y; z 2 S)(max ff (xyz); g min ff (x); f (z); g):
Proof. It is easy.

_q ) fuzzy interior ideal over S if it satis…es
(i) (for all x; y 2 S)(max ff (xy); g min ff (x); f (y); g);
(ii) (for all x; a; z 2 S)(max ff (xaz); g min ff (a); g):
Proof. It is easy.
Theorem 396 Let S be an AG-groupoid and hF; Ai a fuzzy soft set over
S. Then
(i) hF; Ai is an (2 ; 2 _q ) fuzzy soft left ideal (resp., right, bi-ideal,

quasi-ideal) over S if and only if non-empty subset F (")r is a left ideal
(resp. right, bi-ideal, quasi-ideal) of S for all " 2 A and r 2 ( ; ]:
(ii) If 2 = 1 + ; then hF; Ai is an (2 ; 2 _q ) fuzzy soft left ideal
(resp., right, bi-ideal, quasi-ideal) over S if and only if non-empty subset
hF (")ir is a left ideal (resp. right, bi-ideal, quasi-ideal) of S for all " 2 A
and r 2 ( ; 1]:
(iii) hF; Ai is an (2 ; 2 _q ) fuzzy soft left ideal (resp., right, bi-ideal,
quasi-ideal) over S if and only if non-empty subset [F (")]r is a left ideal
(resp. right, bi-ideal, quasi-ideal) of S for all " 2 A and r 2 ( ; minf2
; 1g]:
Proof. It is straightforward.
Corollary 397 Let S be an AG-groupoid and P S: Then P is a left

ideal (resp. right ideal, bi-ideal, quasi-ideal) of S if and only if (P; A) is
an (2 ; 2 _q )-fuzzy soft left ideal (resp., right ideal, bi-ideal, quasi-ideal)
over S for any A E:
10.1 Some Characterizations Using Generalized

Fuzzy Soft Bi-ideals
Theorem 398 For an AG-groupoid with left identity e, the following are
equivalent.
(ii) B = B 2 , for any bi-ideal B.
Proof. It is easy.
Theorem 399 Let S be an AG-groupoid with left identity e. Then S is

intra-regular if and only if hF; Ai \ hG; Bi b( ; ) hF; Ai hG; Bi for any
(2 ; 2 _q )-fuzzy soft bi- ideal hF; Ai and hG; Bi over S.
Proof. Let S be an intra-regular and hF; Ai and hG; Bi are any two (2
; 2 _q )-fuzzy soft bi-ideal of S . Now let x be an element of S, " 2 A [ B
and hF; Ai \ hG; Bi = hH; A [ Bi: We consider the following cases.
Case 1: " 2 A B: Then H(") = F (") = (F G)("):
Case 2: " 2 B A: Then H(") = G(") = (F G)("):
Case 3: " 2 A\B:Then H(") = F (")\G(") and (F G)(") = F (") G("):
Now we show that F (")\G(") b _q( ; ) F (") G("): Since S is intra-regular,
therefore for any a in S there exist x and y in S such that a = (xa2 )y:
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a:

= (y(x((xa2 )y))a = (y((xa2 )(xy)))a = ((xa2 )(y(xy)))a
= ((a(xa))(xy 2 ))a = (((xy 2 )(xa))a)a
= (((xy 2 )(x((xa2 )y)))a)a = (((xy 2 )(xa2 )(xy)))a)a
= ((xa2 )((xy 2 )(xy)))a)a = (((xy)(xy 2 )(a2 x))a)a
= ((a2 (((xy)(xy 2 ))x))a)a
= ((a2 u)a)a, where u = ((xy)(xy 2 ))x
= [fa2 (eu)ga]a = [f(ue)a2 ga]a = [fa((ue)a)ga]a
= [(av)a]a, where v = (ue)a).
Then we have
max f(F (") G("))(a); g
= max sup min fF (")(p); G(")(q)g ;

a=pq
max fmin fF (")((av)a); G(")(a)g ; g
= min fmaxfF (")((av)a); g; maxfG(")(a); gg
min fmin fF (")(a); F (")(a); g; minfG(")(a); gg
= min fmin fF (")(a); g ; min fG(")(a); gg
= min f(F (") \ G("))(a); g :
It follows that F (")\G(") _q( ; ) F (") G("): That is H(") _q( ; ) (F

G)("): Thus in any case, we have
H(") _q( ; ) (F G)("):
Therefore
hF; Ai \ hG; Bi b( ; ) hF; Ai hG; Bi:
Conversely assume that the given condition hold. Let B be any bi-ideal
of S then (B; E) is an (2 ; 2 _q ) fuzzy soft bi-ideal of S. Now by the
assumption, we have (B; E) \ (B; E) b( ; ) (B; E) (B; E). Hence
we have
B = ( ; ) (B\B) =( ; ) B \ B q( ; ) B B
= ( ; ) BB =( ; ) B2 :
It follows that B B 2 : Also B 2 B. This implies that B = B 2 .

Therefore S is intra-regular.
Theorem 400 In intra-regular AG-groupoid S with left identity the fol-

(i) A fuzzy subset f of S is an (2; 2 _qk )-fuzzy right ideal.
(ii) A fuzzy subset f of S is an (2; 2 _qk )-fuzzy left ideal.
(iii) A fuzzy subset f of S is an (2; 2 _qk )-fuzzy bi-ideal.
(iv) A fuzzy subset f of S is an (2; 2 _qk )-fuzzy interior ideal.
(v) A fuzzy subset f of S is an (2; 2 _qk )-fuzzy quasi-ideal.
Proof. It is easy.
ing conditions are equivalent.
(ii) For all left ideals A; B; A \ B BA:
(iii) For all (2; 2 _qk )-fuzzy left ideals f and g, f ^k g g k f:
(iv)For all (2; 2 _qk )-fuzzy bi-ideals f and g, f ^k g g k f:
(v)For all (2; 2 _qk )- generalized fuzzy bi-ideals f and g, f ^k g g k f:
Proof. It is easy.
Theorem 402 Let S be an AG-groupoid with left identity e. Then S is

intra-regular if and only if hG; Ri \ hF; Qi b( ; ) hF; Qi hG; Ri for any
(2 ; 2 _q )-fuzzy soft left ideal hF; Qi and for any (2 ; 2 _q )-fuzzy soft
ideal hG; Ri over S.
Proof. Let S be an intra-regular and hF; Qi and hG; Ri are any (2 ; 2

_q )-fuzzy soft left ideal and (2 ; 2 _q )-fuzzy soft ideal of S . Now let
x be an element of S, " 2 Q [ R and hF; Qi \ hG; Ri = hH; Q [ Ri: We
consider the following cases.
Case 1: " 2 Q R: Then H(") = F (") = (F G)("):
Case 2: " 2 R Q: Then H(") = G(") = (F G)("):
Case 3: " 2 Q \ R:Then H(") = F (") = F (") \ G(") and (F G)(") =
F (") G("): Now we show that F (") \ G(") b( ; ) G(") F ("): Since S is
intra-regular, therefore for any a in S there exist x and y in S such that
a = (xa2 )y:
a = (xa2 )y = (x(aa))y = (a(xa))y = (y(xa))a = (y(xa))(ea)

= (ye)((xa)a) = (xa)((ye)a) = (xa)(ta); where t = (ye):
Then we have
max f(G(") F ("))(a); g
= max sup min fG(")(u); F (")(v)g ;

a=uv
max fmin fG(") (xa) ; F (")(ta)g ; g
= min fmaxfG(")(xa)); g; maxfF (")(ta); gg
min fmin fG(")(a); g; minfF (")(a); gg
= min fmin fG(")(a); F (")(a); gg
= min f(G(") \ F ("))(a); g
= min f(F (") \ G("))(a); g :
It follows that F (")\G(") _q( ; ) G(") F ("): That is H(") _q( ; ) (G

F )("): Thus in any case, we have
H(") _q( ; ) (G F )("):
Therefore,
hF; Qi \ hG; Ri b( ; ) hG; Ri hF; Qi:
Conversely assume that the given condition hold. Let L1 and L2 are any
two left ideal of S then, (L1 ; E) and (L2 ; E) are
(2 ; 2 _q )-fuzzy soft left ideal of S. Now by the assumption, we have
(L1 ; E) \ (L2 ; E) b( ; ) (L2 ; E) (L1 ; E):Hence we have
(L1 \L2 ) = ( ; ) L1 \ L2
q( ; ) L2 L1 =( ; ) L2 L1
It follows that L1 \ L2 L2 L1 : Therefore by theorem S is intra-regular.
10.2 References
[1] S.K. Bhakat and P. Das, On the de…nition of a fuzzy subgroup, Fuzzy
Sets and Systems, 51(1992); 235-241:
[2] S.K. Bhakat and P. Das, (2; 2 _q)-fuzzy subgroups, Fuzzy Sets and
Systems, 80(1996); 359-368:
[3] A.H. Cli¤ord and G.B. Preston, The algebraic theory of semigroups,
Volume II, Amer. Math. Soci., 1967:
[4] K. A. Dib and N. Galham, Fuzzy ideals and fuzzy bi-ideals in fuzzy
semigroups, Fuzzy Sets and systems 92(1992); 103 111; 1992:
[5] B. Davvaz, (2; 2 _q)-fuzzy subnearrings and ideals, Soft Comput.,

10(2006); 206-211:
[6] B. Davvaz and O. Kazanci, A new kind of (2; 2 _q)-fuzzy sublattice
(ideal, …lter) of a lattice, International Journal of Fuzzy Systems, 13(1)
(2011), 55-63.
[7] F. Feng, C.X. Li, B. Davvaz, M.I. Ali, Soft sets combined with fuzzy
sets and rough sets: a tenta approach, Soft Computing 14 (2010), 899–
911.
[8] F. Feng, X.Y. Liu, V. Leoreanu-Fotea, Y.B. Jun, Soft sets and soft
rough sets, Inform. Sci., 181 (2011), 1125–1137.
[9] F. Feng, Y. B. Jun, X. Zhao, Soft semirings, Comput. Math. Appl.
56(2008); 2621 2628.
[10] Florentin Smarandache, (t,i,f)-Neutrosophic Structures and I-
Neutrosophic Structures, Neutrosophic Sets and Systems, 3-10, Vol.
8, 2015.
[11] Florentin Smarandache, n-Valued Re…ned Neutrosophic Logic and Its
Applications in Physics, Progress in Physics, 143-146, Vol. 4, 2013.
[12] I.R. Goodman, Fuzzy sets as equivalence classes of random sets, in:
Recent Developments in Fuzzy Sets and Possibility Theory (R. Yager,
Ed.), Pergamon, New York (1982).
[13] P. Holgate, Groupoids satisfying a simple invertive law, The Math.
Stud., 1 4; 61 ; 101 106(1992).
[14] J. M. Howie, An Introduction to semigroup theory, Academic Press
London; 1976:
[15] J. M. Howie, The maximum idempotent separating congruence on an
inverse semigroup, Proceedings of the Edinburgh Mathematical Soci-
ety, 06 (1964), 71-79.
[16] M. Irfan Ali, F. Feng, X. Liu, W. K. Min, M. Shabir, On some new
operations in soft set theory, Comput. Math. Appl.57(2009)1547
1553.
[17] K. Iseki, A characterization of regular semigroups, Proc. Japan Acad.,
32 (1965), 676 677.
[18] Y. B. Jun, Soft BCK/BCI-algebras, Comput. Math Appl.
56(2008)1408 1413.
[19] Y.B. Jun and S.Z. Song, Generalized fuzzy interior ideals in semi-
groups, Inform. Sci., 176(2006); 3079-3093:
[20] J. Jeµzek and T. Kepka, A note on medial division groupoids, Proc.

Amer. Math. Soc., 2; 119(1993); 423 426:
[21] J. Jeµzek and T. Kepka, The equational theory of paramedial cancella-

tion groupoids, Czech. Math. J., 50; 125(2000); 25 34:
[22] Y. B. Jun, Generalizations of (2; 2 _q)-fuzzy subalgebras in

BCK/BCI-algebra, Comput. Math. Appl., 58 (2009) ; 1383-1390.
[23] Faiz Muhammad Khan, Asghar Khan, and Nor Haniza Sarmin, Char-
acterizations of Ordered Semigroups by (2 ; 2 _q )-Fuzzy Interior
Ideals,.ISSN1995-0802, Lobachevskii Journal of Mathematics, 2011,
Vol. 32, No. 4, pp. 278 288:
[24] M. A. Kazim and M. Naseeruddin, On almost semigroups, The Alig.

Bull. Math., 2 ; 1 7(1972):
[25] O. Kazanc¬and S. Yamak, Generalized fuzzy bi-ideals of semigroups,

Soft Comput., 12(2008), 1119 1124.
[26] N. Kehayopulu and M. Tsingelis, Regular ordered semigroups in terms

of fuzzy subsets, Inform. Sci. 176 (2006) 3675-3693.
[27] N. Kuroki, Fuzzy bi-ideals in semigroups, Comment. Math. Univ. St.

Pauli, 27 (1979), 17-21.
[28] N. Kuroki, Fuzzy semiprime quasi ideals in semigroups, Inform. Sci.,

75(3)(1993)201 211:
[29] Madad Khan and Nouman A. Khan, On fuzzy Abel-Grassmann’s

groupoids, Advances in Fuzzy Mathematics,Vol 5; No 3(2010); 349
360:
[30] Madad Khan, Florentin Smarandache, Tariq Aziz, Fuzzy Abel Grass-
mann Groupoids […rst version], by Educational Publisher, Columbus,
82 p., 2015; [second updated and enlarged version], 110 p., 2015.
[31] S. Lajos, A note on semilattice of groups, Acta Sci. Math. Szeged,

31 (1982) ; 179 180.
[32] D. Molodtsov, Soft set theory–First results, Comput. Math. Appl. 37

(1999) 19–31.
[33] Molodstov, Soft set theory-First results, Comput. Math . Appl.37(4

5)(1999)19 31.
[34] Madad Khan and Tauseef Asif, Characterizations of regular left almost
semigroups by their fuzzy ideals, Journal of Advanced Research in
Dynamical and Control Systems, Vol 2, No 2 (2010), 59-75
[35] P. K. Maji, R. Biswas, A. R. Roy, Fuzzy soft sets, J. Fuzzy

Math:9 (3 )(2001 )589 602 :
[36] Madad Khan, Faisal and Venus Amjid, On some classes of Abel-
Grassmann’s groupoids, Journal of Advanced Research in Pure Math-
ematics, 3, 4 (2011), 109-119.
[37] J. N. Mordeson, D. S. Malik and N. Kuroki, Fuzzy semigroups.

Springer-Verlag, Berlin, Germany (2003).
[38] Madad Khan and N. Ahmad, 2010, Characterizations of left almost

semigroups by their ideals, Journal of Advanced Research in Pure
Mathematics, 2; 61 73:
[39] Q. Mushtaq and S. M. Yusuf, 1988, On LA-semigroup de…ned by a

commutative inverse semigroups, Math. Bech., 40, pp. 59 62.
[40] Q. Mushtaq and Q. Iqbal, Decomposition of a locally associative LA-

semigroup, Semigroup Forum, 41(1990); 155 164:
[41] Q. Mushtaq and Madad Khan, Semilattice decomposition of locally

associative AG -groupoids, Algeb. Colloq., 16, 1(2009), 17 22.
[42] V. Murali, Fuzzy points of equivalent fuzzy subsets, Inform. Sci.,

158(2004); 277-288.
[43] Q. Mushtaq and S. M. Yusuf, On LA-semigroupss, The Alig. Bull.

Math., 8 (1978); 65 70:
[44] J. Maiers and Y. S. Sherif, Applications of fuzzy set theory, IEEE

Transactions on Systems, Man and Cybernetics, 15(1) (1985), 175-
189.
[45] M. Petrich, Introduction to Semigroups, Charles.E. Merrill, Columbus,

1973.
[46] P. V. Protić and N. Stevanović, On Abel-Grassmann’s groupoids, Proc.

Math. Conf. Priština, (1994); 31 38:
[47] P. V. Protić and M. Boµzinović, Some congruences on an

AG -groupoid, Algebra Logic and Discrete Mathematics, Niš,
3; 9(1995); 879 886:
[48] P. V. Protić and N. Stevanović, AG-test and some general properties

of Abel-Grassmann’s groupoids, PU. M. A., 4(6)(1995), 371 383.
[49] P.M Pu and Y.M. Liu, Fuzzy topology I, neighborhood structure of

a fuzzy point and Moore-Smith convergence, J. Math. Anal. Appl.,
76(1980); 571-599.
[50] G. Pilz, Near rings, The theory and its applications, North Holland
Publishing Company Amsterdam, Revised edition, 1983.
[51] P.V. Protić and Stevanović, AG-test and some general properties of
Abel-Grassmann’s groupoids, PU. M. A., 4, 6 , 371 383(1995).
[52] Z. Pawlak, Rough sets, Int. J. of Inform. Comput. Sci. 11 (1982) 341-
356.
[53] A. Rosenfeld, Fuzzy groups, J. Math. Anal. Appl., 35 (1971), 512 517.
[54] A.R. Roy, P.K. Maji, A fuzzy soft set theoretic approach to decision
making problems, Journal of Computational and Applied Mathematics
203 (2007) 412-418.
[55] M. Shabir, Y.B. Jun and Y. Nawaz, Semigroups characterized by (2
; 2 _qk )-fuzzy ideals, Comput. Math. Appl., 60 (2010), 1473-1493.
[56] M. Shabir, Y.B. Jun and Y. Nawaz, Characterizations of regular semi-
groups by ( ; )-fuzzy ideals, Comput. Math. Appl., 59 (2010), 161-
175.
[57] O. Steinfeld, Uber die Quasi-ideal von Halbgruppen, Publ. Math. De-
brece., 4 (1956) ; 262 275.
[58] W. B. Vasantha Kandasamy, Florentin Smarandache, Ilanthenral K.,
Distance in Matrices and Their Applications to Fuzzy Models and
Neutrosophic Models, 169 p., EuropaNova, Brussels, Belgium, 2014.
[59] W. B. Vasantha Kandasamy, Florentin Smarandache, Ilanthenral K,
New Techniques to Analyse the Prediction of Fuzzy Models, Eu-
ropaNova, Brussels, Belgium, 242 p., 2014.
[60] W. B. Vasantha Kandasamy, Florentin Smarandache, Ilanthenral K,
Pseudo Lattice Graphs and their Applications to Fuzzy and Neutro-
sophic Models, EuropaNova, Brussels, Belgium, 275 p., 2014.
[61] W. B. Vasantha Kandasamy and F. Smarandache, Algebraic Struc-
tures on the Fuzzy Interval [0, 1), 250 p., Educational Publisher, 2014.
[62] W. B. Vasantha Kandasamy, Florentin Smarandache, Fuzzy Neutro-
sophic Models for Social Scientists, Education Publisher, Columbus,
OH, 167 pp., 2013.
[63] W. B. Vasantha Kandasamy, Smarandache Fuzzy Algebra, Bookman
Publishing, 2003.
[64] W. B. Vasantha Kandasamy, Florentin Smarandache, Fuzzy Cognitive
Maps and Neutrosophic Cognitive Maps, Xiquan, Phoenix, 211 p.,
2003
[65] Y. Yin and J. Zhan, Characterization of ordered semigroups in terms

of fuzzy soft ideals, Bull. Malays. Math. Sci. Soc., (2) 35(4) (2012),
997-1015.
[66] J. Zhan and Y.B. Jun, Generalized fuzzy interior ideals of semigroups,
Neural Comput. Applic., 19(2010); 515 519.
[67] L. A. Zadeh, Fuzzy sets. Inform. Control, 8 (1965), 338 353.
An AG-groupoid is an algebraic structure that lies in between a groupoid
and a commutative semigroup. It has many characteristics similar to that of a
commutative semigroup. If we consider x2y2= y2x2, which holds for all x, y in a
commutative semigroup, on the other hand one can easily see that it holds in an
AG-groupoid with left identity e and in AG**-groupoids. This simply gives that
how an AG-groupoid has closed connections with commutative agebras.
We extend now for the first time the AG-groupoid to the Neutrosophic
AG-groupoid. A neutrosophic AG-groupoid is a neutrosophic algebraic
structure that lies between a neutrosophic groupoid and a neutrosophic
commutative semigroup.

Theory of Abel Grassmann's Groupoids

Uploaded by

Copyright:

Available Formats

Theory of Abel Grassmann's Groupoids

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Theory of Abel Grassmann's Groupoids

Uploaded by

Copyright:

Available Formats

Educational Publisher

Prof. Rajesh Singh, School of Statistics, DAVV, Indore (M.P.), India.

Dr. Linfan Mao, Academy of Mathematics and Systems,

Mumtaz Ali, Department of Mathematics, Quaid-i-Azam University,

Said Broumi, University of Hassan II Mohammedia,

Theory of Abel Grassmann's Groupoids

Copyright: © Publisher, Madad Khan1, Florentin Smarandache2, Saima Anis1. 2015

The Educational Publisher, Inc.

1 Congruences on Inverse AG-groupoids 11

2 Structural Properties of -AG -groupoids 21

3 Embedding and Direct Product of AG-groupoids 47

4 Ideals in Abel-Grassmann’s Groupoids 61

5 Some Characterizations of Strongly Regular AG-groupoids 101

6 Fuzzy Ideals in Abel-Grassmann's Groupoids 109

7 (2; 2 _q) and (2; 2 _qk)-fuzzy Bi-ideals of AG-groupoids 119

8 Interval Valued Fuzzy Ideals of AG-groupoids 151

10 On Fuzzy Soft Intra-regular Abel-Grassmann's Groupoids 191

10.1 Some Characterizations Using Generalized Fuzzy Soft Bi-ideals 199

practical applications. The role of fuzzy theory in automata and formal

[(a1 + b1 I) (a2 + b2 I)] (a3 + b3 I) = [a1 a2 + b1 b2 I] (a3 + b3 I)

[(a3 + b3 I) (a2 + b2 I)] (a1 + b1 I) = [a3 a2 + b3 b2 I] (a1 + b1 I)

In chapter one we discuss congruences in an AG-groupoid. In this chapter

a(bc) = b(ac); for all a; b; c 2 S: (1)

Note that a commutative AG-groupoid S with left identity becomes a

(ab)c = (cb)a = a(cb) = a(bc).

In [15] J. M. Howie de…ned a relation as (a; b) 2 if and only if

Example 1 Let S = f1; 2; 3g and the binary operation “ ” de…ned on S as

Lemma 2 Let S be an inverse AG -groupoid and de…ned by a b, if

Proof. Clearly is re‡exive, symmetric and transitive, so is an equiva-

Lemma 3 Let S be an inverse AG -groupoid, then the relation = f(a; b) 2

Proof. It is available in [47].

Lemma 4 The congruence relation is equivalent to :

1.2 Inverse AG -groupoids

Theorem 8 Let E be the set of all idempotents of S and let be a congru-

Thus ac min bc. Therefore min is a congruence relation.

E = fz 2 S : ez = ze for every e in Eg:

Proof. Let S be an inverse AG -groupoid and let be the idempotent

ea = e((aa 1 )a) = (aa 1 )(ea) = ((ea)a 1 )a

Similarly we can show that b 1 b = (a 1 a)(b 1 b). Therefore a 1 a = b 1 b:

Theorem 13 Let E be the set of all idempotents of S and let be the

Proof. Since is idempotent separating congruence so S= is a semilattice

but this holds for any x in S; so E = S:

Theorem 14 Let E be the set of all idempotents of S and let be the

Proof. Let = 1S ; Then for z 2 E implies that ze = ez; for all e 2 E if

Therefore zf 1 2 E : Then by theorem 5, z zz 1 ; but = 1S ; so z =

= ((xy 1 )y)x 1 = ((x 1 y)y)x 1

Theorem 15 Let E be the set of all idempotents of S then the relation

Lemma 17 Let E be the set of all idempotents of S then the centralizer

Proof. Let a; b 2 E ; then ae = ea and be = eb; for all e 2 E, so

2.1 Gamma Ideals in -AG-groupoids

A -AG -groupoid also satis…es the -paramedial law for all w; x; y; z 2

De…nition 21 A -AG-subgroupoid A of a -AG-groupoid S is called a

Let us de…ne = f , , g as follows.

Clearly S is an intra-regular -AG-groupoid because 1 = (2 (1 1)) 3;

a 2 S = (S S) S = ((S a) (S a)) S = ((S S) (a a)) S

Which shows that S is an intra-regular -AG -groupoid.

b = (x (b b)) y = (b (x b)) y = (y (x b)) b

Which shows that (B S) B = B \ S:

i = (x (i i)) y = (i (x i)) y = (y (x i)) i

Which shows that (S I) S = S \ I: