Communication, Acting and Perceiving: Artificial Intelligence
Communication, Acting and Perceiving: Artificial Intelligence
Communication, Acting and Perceiving: Artificial Intelligence
Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Artificial Intelligence
Communication, Acting
and Perceiving
Submitted By:
Nene, Senny M.
Del Rosario Genesis D.C
Francisco, Jasper
Beroy, Justine
Flores, Weastlie
Submitted To:
Mr. Brayan Acebuche
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Communication
Communication is the intentional exchange of information brought about by the
production and perception of signs drawn from a shared system of conventional signs.
What sets humans apart from other animals is the complex system of structured
messages known as language that enables us to communicate most of what we know
about the world. Although chimpanzees, dolphins, and other mammals have shown
vocabularies of hundreds of signs and some aptitude for stringing them together, only
human can reliably communicate an unbounded number of qualitative different
messages.
Communication as Action
1. Speech act is one of the actions available to an agent to produce language.
2. Speaker, Hearer and Utterance terms of referring any mode of communications.
3. Word is to refer any kind of conventional communicative sign.
4. Indirect speech act that agents in multi agent environment can use
communication to help arrive at joint plans. For example, a group of agents
exploring the wumpus world together gains an advantage (collectively and
individually) by being able to do the following:
Query:This is typically done by asking questions. Have you smelled the wumpus
anywhere?
Inform:this is done by making representative statements: There’s a breeze here
in 3 4
Request: other agents to perform actions: Please help me carry the gold.
Sometimes an indirect speech act(a request in the form of a statement or
question) I could use some help carrying this. An agent with authority can give
commands (Alpha go right; Bravo and Charlie go left), and an agent with power
can make a threat (Give me the gold, or else). These kinds of speech acts are
called directives.
Acknowledge request: OK
Promise or commit to a plan: I’ll shoot the wumpus; you grab the gold.
5. Declarative is another class of speech act that appears to have a more direct
effective on the world, as in I now pronounce you man and wife or Strike three
you’re out. The effect is achieved by creating or confirming a complex web of
mental states among the agents involved.
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
2. Generation- The speaker plans how to turn the proposition P into an utterance
that will make it likely that the hearer, upon perceiving the utterance in the current
situation , can infer the meaning P (or something close to it). Assume that the
speaker is able to come up with the words “The wumpus is dead,” and call this
W.
3. Synthesis- the speaker produces the physical realization W1 of the words W.
4. Perception- H perceives the physical realization W1 as W2 and decodes it as
the words W2.
5. Analysis- H infers that W2 has possible meaning P1,…P2 Analysis has three
main parts: syntactic interpretation(or parsing), semantic interpretation, and
fragmantic interpretation.
Parsing- is the process of building parse tree for an input string. The interior
nodes of the parse tree represent phrases and the leaf nodes represent words.
Semantic interpretation- is the process of extracting the meaning of an utterance
as an expression in some representation language.
Pragmatic Interpretation- takes into account the fact that some words can have
different meanings in different situation.
6. Disambiguation- H infers that S intended to convey Pi (where ideally Pi = P). Most
speaker are not intentionally ambiguous, but most utterances have feasible
interpretation.
7. Incorporation- H decides to believe Pi (or not). A totally naïve agent might believe
everything it hears, but a sophisticated agent treats the speech act as evidence for Pi,
not confirmation of it.
A Formal Grammar for a Fragment of English
The Lexicon of
First we define the lexicon, or list allowable words. The words are grouped into
categories or parts of speech familiar to dictionary users: nouns, pronouns, and names
to denote things, verbs to denote events, adjectives to modify nouns, and adverbs to
modify verbs. Categories that are perhaps less familiar to some readers are articles
(such as the), prepositions (in), and conjunctions (and).Noun, Verbs, adjective and
adverbs these four categories are called open classes Pronoun, article, preposition and
conjunction are called closed classes.
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
The next step is to combine the words into phrases. We will use the five nonterminal
symbols to define the different kinds of phrases: sentence (S), noun phrase (NP),
verb phrase (VP), prepositional phrase (NP), and relative clause (RelClause).
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Semantic Interpretation
• is an important component in dialog systems
• It is related to natural language understanding, but mostly its refers to the last
stage of understanding.
• The goal of interpretation is binding the user utterance to concept, or something
the system can understand.
• Semantics – The extraction of the meaning of utterances
• Semantic interpretation is the process of associating expression with a phrase
• The meaning of phrase “The wumpus” is the big, hairy beast that represent in
logic as the logical term Wumpus.
• And the meaning of “the wumpus is dead” is the logical sentence
dead(Wumpus)
Parse tree with semantic interpretation
• 3 + (4 / 2 )
Times and Tense
• We want to represent the difference between “john loves mary” and “john loved
mary”
• English uses verb tenses(past, present, and future) to indicate the relative time of
an event.
Quantification
• Consider the sentence “Every agent smells a wumpus”.
• The sentence is actually ambiguous the preferred meaning is that the agents
might be smelling different wumpus
• But an alternative meaning is that there is a single wumpus that everyone smells.
Intermediate form
• Many modern grammars take a different tack. They define an intermediate form
to mediate between syntax and semantics.
• The intermediate form has two key properties
• First, it is structurally similar to the syntax of the sentence and thus can be easily
constructed through compositional means.
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
This approach works for the simple grammars in this chapter but there can be
difficulties in scaling up to larger grammar. The search strategy used by logical
inference engine is important depth-first strategies can lead to infinite loops.
Long-distance dependencies – between constituent. For example in “Whom” did
the agent tell you to give the gold to ? The final word “to” should be parsed as a
PP with a missing a NP the missing NP is licensed by the first word of the
sentence “who” A complex system of augmentations is used to make sure that
the missing NPs match up with the licensing words.
AMBUIGUITY AND DISAMBIGUATION
In some cases hearers are consciously aware of ambiguity in an utterance.
But most of the time the language we hear seems unambiguous. Thus, when
researchers first began to use computers to analyze language in the 1960s they
were quite surprised to learn that almost every utterance is highly ambiguous,
even though the alternative interpretations might not be apparent to a native
narrative speaker. A system with a large grammar and lexicon might find
thousands of interpretations for a perfectly ordinary sentence.
Disambiguation
Is the process of resolving conflicts that arise when a potential article title
is ambiguous, most often because it refers to more than one subject covered by
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
The Language model: the likelihood that a certain string of words will be chosen
given that the speaker has the intention of communicating a certain fact.
The acoustic model: the likelihood that a particular sequence of sounds will be
generated given that the speaker has chosen a given string of words.
The structure of coherent discourse
Enable or Cause
S1 brings about a change of state (which may be implicit) that causes or enable
s2. Example “I went outside. I drove to school. “(Going outside enables the
implicit getting into a car.
Explanation: The reverse of enablement: s2 causes or enables s1 and thus is an
explanation for it example: “ It was late for school. I overslept.
Ground Figure: s1 describe a setting or background for s2 Example: “ It was a
dark and stormy night . Rest of story”.
Evaluation: From s2 infer that s1 is part of the speaker s plan for executing the
segment as a speech act. Example: “ A funny thing happened . Rest of story.
Exemplification: S2 is an example of the general principle in S1. Example “ This
algorithm reverses a list. The input {A, B, C] is mapped to { C, B, A}.
Focus Spaces
Certain utterances cause the focus to shift by pushing or popping elements of the
stack. For example in the restaurant story the sentence john went to a fancy
restaurant pushes new focus on to stack. Within that focus the speaker can use a
define a NP to refer the waiter if the story continued with “ John went home then
the focus space would be popped from the stack and discourse could no longer
refer to the waiter with the waiter or he.
Generalization: S1 is an example of the general principle in S2 example: “{A, B,
C} is mapped to {C, B, A}. In general reverses algorithm list.
Violated Expectations: Infer – P from S2, negating the normal interface of P from
S1 Example “ This paper is weak. On the other hand it is interesting.
A discourse has structure above the level of a sentence. We can examine this
structure with the help of a grammar of discourse.
Segment(x) S(x)
Segment( Coherence Relation)(x,y)) Segment(x) Segment(y)
This grammar says that a discourse is composed of segments where each
segments is either a sentence or a group of sentences and where segments are
joined by coherence relations.
Sta. Cecilia College
108 Gen. T. De Leon, Val. City
Tel. No. 293-60-15 to 17/ Telefax: 293-60-16
Website: www.santacecilia.com.ph
Grammar Induction
also known as grammatical inference or syntactic pattern recognition, refers to
the process in machine learning of learning a formal grammar (usually as a
collection of re-write rules or productions or alternatively as a finite state
machine or automaton of some kind) from a set of observations, thus
constructing a model which accounts for the characteristics of the observed
objects.
Grammatical inference has often been very focused on the problem of learning
finite state machines of various types (see the article Induction of regular
languages for details on these approaches), since there have been efficient
algorithms for this problem since the 1980s.
Discourse
As discourse, an enouncement (statement) is not a unit of semiotic signs, but an
abstract construct that allows the semiotic signs to assign meaning, and so
communicate specific, repeatable communications to, between, and among
objects, subjects, and statements. Therefore, a discourse is composed of
semiotic sequences (relations among signs that communicate meaning) between
and among objects, subjects, and statements.