Doctus Documentation

Doctus Documentation
Introduction
1 Foreword ................................................................................................................................. 3
2 About This Document ............................................................................................................ 3
3 List of Figures and Tables ...................................................................................................... 5
3.1 List of Figures .......................................................................................................................................... 5
3.2 List of Tables ............................................................................................................................................ 7
Doctus Guide
1 Doctus KBS ............................................................................................................................. 8
2 Original Decision (when there is no experience in the domain) ........................................... 9
2.1 Benefits...................................................................................................................................................... 9
2.2 Acquisition of Attributes ........................................................................................................................ 9
2.3 Hierarchy of the Attributes: the Rule-Based Graph ......................................................................... 10
2.4 Acquisition of Cases .............................................................................................................................. 11
2.5 Knowledge Import Data Mining ..................................................................................................... 11
2.6 The Rules ................................................................................................................................................ 13
2.7 The Reasoning........................................................................................................................................ 16
2.8 Decision Analyses and Fine-Tuning ................................................................................................... 17
2.9 Knowledge Export Intelligent Portal .............................................................................................. 18
3 Routine Decision (when there are few dozen cases in the domain) ..................................... 19
3.1 Benefits.................................................................................................................................................... 20
3.2 Acquisition of Attributes ...................................................................................................................... 20
3.3 Acquisition of Cases .............................................................................................................................. 21
3.4 Knowledge Import Data Mining ..................................................................................................... 21
3.5 Decision Tree: the Case-Based Graph ............................................................................................... 23
3.6 Classification of Cases........................................................................................................................... 24
3.7 The Reasoning........................................................................................................................................ 25
3.8 Decision Analyses and Fine-Tuning ................................................................................................... 25
4 Learning from Cases (reduction of the model) .................................................................... 28
4.1 Benefits.................................................................................................................................................... 28
4.2 Single-Level Hierarchy .......................................................................................................................... 29
4.3 Missing or Indefinite Rules .................................................................................................................. 29
4.4 The Reasoning........................................................................................................................................ 30
4.5 Tacit Knowledge and Fine-Tuning ..................................................................................................... 30
Doctus Reference
1 A Knowledge-Based System ................................................................................................. 33
2 The Five Panes of Doctus ..................................................................................................... 33
2.1 Attributes ................................................................................................................................................ 33
2.2 Cases ........................................................................................................................................................ 41
2.3 Rule-Based Graph ................................................................................................................................. 47
2.4 Rules ........................................................................................................................................................ 50
2.5 Case-Based Graph ................................................................................................................................. 59
3 Navigation in Tables............................................................................................................. 67
3.1 Moving Cursor ....................................................................................................................................... 67
3.2 Context Menu ........................................................................................................................................ 67
3.3 Entering Text ......................................................................................................................................... 67
3.4 Selection .................................................................................................................................................. 67
3.5 Inserting and Deleting .......................................................................................................................... 67
3.6 Moving Cells........................................................................................................................................... 67
3.7 Sizing Cells .............................................................................................................................................. 68
4 Knowledge Export & Import ................................................................................................ 68
4.1 Knowledge Export ................................................................................................................................ 68
4.2 Knowledge Import ................................................................................................................................ 71
5 User Controls ........................................................................................................................ 76
5.1 Main Menu.............................................................................................................................................. 76
5.2 Keyboard Shortcuts............................................................................................................................... 84
5.3 Toolbar .................................................................................................................................................... 85
6 Operating & Maintenance .................................................................................................... 85
6.1 Installation .............................................................................................................................................. 85
6.2 Demo and Collector modes ................................................................................................................. 86
6.3 Doctus File Types.................................................................................................................................. 86
6.4 Advanced Mode ..................................................................................................................................... 87
6.5 Starting Doctus ...................................................................................................................................... 87
6.6 Command Line Arguments.................................................................................................................. 88
6.7 Uninstall .................................................................................................................................................. 88
7 Appendix ............................................................................................................................... 88
7.1 Samples ................................................................................................................................................... 88
7.2 Export Templates .................................................................................................................................. 88
7.3 Batch Commands .................................................................................................................................. 92
7.4 Determining Informativity ................................................................................................................... 92
7.5 Glossary .................................................................................................................................................. 93
Introduction http://www.doctus.info
Introduction
1 Foreword
by Zoltn Baracskai
In the seventies were great prospects in the applications of Artificial

Intelligence; which deflated in the eighties. Back then I wrote a book about the
adolescence of Expert Systems. Rereading the text now, I see how much one
can know by only reading about things, without actually trying. How much
better it would have been if I had had the experience of business applications.
Only the creations of some informaticians were available who were people
without any business background.
In the last 20 years I developed my own system and I have about a hundred of running applications.
Now I am able to unfold the real possibilities of expert system applications in business. And I have no
reason to deprive the new generation of things what I have been curtailed of.
From the many years of experience with Expert System Shells a couple of principles rose, that are
worth to consider:
When facing a new solution, you should check whether there is a trial version with business
examples and with a guide and/or help feature just to see in few minutes whether it fits
you.
On the input side, a KBS Shell has to provide the integrability with databases/data
warehouses; while on the output side it should be able to export the knowledge base, in a
convenient form for presentation, e.g. as an html page.
Any action of the reasoning must be transparent and in one click, essentially, it is to be
understandable.
In addition to the software, there is a need for a Knowledge Engineer, whom we can trust,
and whose role is not only to show which button to press, but also how to manage the
acquired knowledge. The consultancy record can already help to build trust.
I paid special attention that the members of my development team follow these principles while
developing the version 3.0 of Doctus KBS Shell.
2 About This Document
This document is provided in three forms: as helpfile included into Doctus KBS, as online guide
available from different Doctus websites (http://www.doctus.info, http://www.doctus.hu,
http://www.odluka.com) and as a printable document file. Since they are all generated from a single
source their contents are identical apart from the special elements, like keywords (only in the help file)
or navigator (in the online version and in the help file). There are two major parts of the document, the
I-3
Doctus Guide and the Doctus References. The first guides the user what to use Doctus for, with focus
on benefits. The second part shows how Doctus is to be used, with focus on the software. To make the
use of the document easier some special notes are distingushed from the body text.
I-4
Tip: Where appropriate we give you tips, based on our experience in consultancy with
Doctus. These tips can be used by everyone who builds a knowledge base.
Maths: Logical and/or mathematical interpretations of features of Doctus, which may be

useful for those interested in deeper understanding.
Technical: Details of technical solutions of Doctus, excluded from regular explanations.

Useful only for users with advanced understanding of programming.
Advanced: Some subtle features and usages of Doctus, which may confuse novices lacking a
deep understanding, are explained. Useful only to those experts with some experience of using
Doctus.
Example: To achive better understanding a number of examples are provided. These are
showed on demo-sized knowledge bases built with Doctus and they are also available for
download under Appendix Samples (R-7.1).
3 List of Figures and Tables
3.1 List of Figures
Figure G-1 Building a Knowledge Base

Figure G-2 Acquiring Attributes and Values.
Figure G-3 The Rule-Based Graph.
Figure G-4 Acquisition of Cases.
Figure G-5 Knowledge Import.
Figure G-6 Clustering External Data.
Figure G-7 Rules in 1D.
Figure G-8 Rules in 2D.
Figure G-9 Advices for Input of Rules.
Figure G-10 Checking the Consistency of the Rule Set.
Figure G-11 Results of Rule-Based Reasoning.
Figure G-12 Explaining the Results.
Figure G-13 Knowledge Export.
Figure G-14 Evaluation in Intelligent Portal.
Figure G-15 Acquiring Attributes and Values.
Figure G-16 Acquisition of Cases.
Figure G-17 Knowledge Import.
Figure G-18 Clustering External Data.
Figure G-19 The Case-Based Graph.
Figure G-20 Informativity and Density of the Attributes.
Figure G-22 Exported Case-Based Knowledge Base used for case collection and evaluation.
Figure G-23 The Case-Based Rule Graph.
Figure G-24 Indefinite Rules in Reduction.
Figure G-26 The exported reduced model.
I-5
Figure R-1 Competitors can be classified based on the amount of their done works into
four categories. They are none, some, mean and many.
Figure R-2 Customize Attributes pane
Figure R-3 Attribute properties Value Ordering
Figure R-4 Attribute properties Branching Method
Figure R-5 Attribute properties Auto Values
Figure R-6 Attribute properties Cluster Configuration
Figure R-7 Attribute properties External Input
Figure R-8 The competitor 1 has some done works. This is vertical view.
Figure R-9 Customize Cases pane
Figure R-10 This case feature is set to some and mean in the rate of 1/3 to 2/3.
Figure R-11 Searching for the competitors of which Tender is fair at least.
Figure R-12 Tender depends on Reference and Finance, Finance depends on Warranty
and Price.
Figure R-13 Customize Rule-Based Graph pane
Figure R-14 Customize Rules pane
Figure R-15 The Rule List view of a rule set with five rules.
Figure R-16 The illustration of the rule set above.
Figure R-17 The same rule set as above in 2D view.
Figure R-18 The illustration of the above rule set. It is the same, isnt it?
Figure R-19 Selecting the domain of the purple rule on Figure R-17.
Figure R-20 The effect of Divide Range on all factors and rules of Figure R-21.
Figure R-21 The effect of Assemble Rules on Figure R-20.
Figure R-22 The effect of Extend Value Ranges on Figure R-21.
Figure R-23 The rule set of Figure R-16 after that its rules are divided. The result is seen on
Figure R-22.
Figure R-24 Selecting rules on the domain of mean Finance and with the output
suggested Tender. Three cells will be selected on Figure R-17.
Figure R-25 Attributes of a knowledge base about the ability of employees. It has no rules
on the Rules pane, but it has well described cases.
Figure R-26 The Case-Based Graph that is generated from the cases of Figure R-25.
Figure R-27 Customize Case-Based Graph pane
Figure R-28 Attributes that can be chosen for branching.
Figure R-29 Cases of the node.
Figure R-30 The distribution of the benchmark case features.
Figure R-31 Generate Case-Based Graph Branching Method
Figure R-32 Generate Case-Based Graph Benchmark
Figure R-33 Generate Case-Based Graph Thresholds
Figure R-34 Generate Case-Based Graph Learning Cases
Figure R-35 The Case-Based Rule Graph created from Figure R-26.
Figure R-36 Exporting into HTML pages
Figure R-37 Exporting pictures of graphs
Figure R-38 Exporting tables of Doctus
Figure R-39 Exporting by an export template
Figure R-40 Importing from an Excel Workbook
Figure R-41 Importing from a database query by Microsoft Query
Figure R-42 Importing from an ODBC database query
Figure R-43 Importing from a mailbox
Figure R-44 Importing from a file with URL Encoded Cases
Figure R-45 Linking the table column done_works to the attribute Done works.
Figure R-46 Font Settings
Figure R-47 Organize Display Styles
Figure R-48 Entering authorization key
I-6
Figure R-49 Choosing running mode
3.2 List of Tables
Table R-1 Keyboard Shortcuts

Table R-2 The Toolbar of Doctus
Table R-3 Parameter Types of Export Templates
Table R-4 Export Templates Desktop Applications
Table R-5 Export Templates Reports
Table R-6 Export Templates Web Clients
Table R-7 Export Templates Web Servers
Table R-8 Batch Commands
I-7
Doctus Guide http://www.doctus.info
Doctus Guide
1 Doctus KBS
Doctus is a Knowledge-Based Expert System Shell. Knowledge-Based Systems support our work
with Knowledge Bases, and the term Expert indicates that this knowledge is on expert level, which
means the highest level of expertise where the knowledge may be made explicit and also signifies a
narrow domain. A Knowledge-Based Expert System consists of two major parts, the Knowledge Base
and the Shell. Being a Shell means that Doctus is an empty software, designed to build the knowledge
bases of the experts, which involves its systematization and, not rarely, discovery of new knowledge.
Building a knowledge base incorporates three processes: Knowledge Acquisition, Knowledge Engineering,
which consist of systematization and fine tuning, and Application, all facilitated by Knowledge Engineer.
(See Figure G-1)
Figure G-1: Building a Knowledge Base
To represent knowledge Doctus uses symbolic logic, that is to say a formalism in which knowledge
is expressed by logical statements consisting of symbols, namely self-defined terms of the expert (i.e.
words) connected by if then rules, also called production rules. Therefore Doctus belongs to
domain of Symbolic Artificial Intelligence.
Knowledge-Based Systems are usually used to evaluate decision alternatives; therefore Doctus
belongs to Decision Support Systems as well. Decision alternatives in Doctus are called cases. The
evaluation of cases is called reasoning. There are three types of reasoning in Doctus:
If the expert can articulate the important aspects of the decision as well as the rules, the
system will trigger these rules to get the evaluation. This is called deduction or Rule-Based
Reasoning. It is used when there is no experience in the domain, therefore the situation calls
for Original Decision (G-2).
If the expert can articulate the aspects but he cannot say which of them are important and he
cannot articulate the rules, though he is experienced enough (a few dozen cases with
evaluation) this experience can be used to find out the rules describing the cases of his
experience using induction, which is the symbolic version of Case-Based Reasoning. As there
is extensive experience in the domain, the situation is described as Routine Decision (G-3).
G-8
From the result of induction the important aspects of the decision can be determined using
reduction. This is the third type of reasoning, though as it can only follow the induction, there
is no third type of knowledge base, only two sorts of knowledge bases are built: rule-based
knowledge base and case-based knowledge base.
2 Original Decision (when there is no experience in the domain)
When there is no experience in the domain the expert is to define the rules, therefore reasoning in
this situation is called Rule-Based Reasoning. As it is started from the generalized rules, which are later
applied to particular cases, it is also called deduction.
2.1 Benefits
If the knowledge engineering was successful, the knowledge-based system would propose the same
evaluation of cases that the expert would. So what is added? The knowledge base is a transparent
description of the knowledge of the expert (or group of experts), which is appropriate to argument the
decision proposal easily.
Apart from the transparency the expert may discover new knowledge, realising that some attributes
were irrelevant, or reshaping his knowledge by understanding the complex rules. This means, that some
tacit relations between the explicit expectations of the expert became explicit.
2.2 Acquisition of Attributes
Knowledge Acquisition always starts with formulation of the aspects of the decision. Aspects are
given by the expert as attributes (i.e. the names of the attributes) and their values. A value of an
attribute is a decision criterion.
Tip: Use short expressions of the special language of your domain of expertise; for better
understanding descriptions may be attached to each of the attributes and values.
The acquisition of attributes and their values happens on the first pane of Doctus named
Attributes. (See Figure G-2)
G-9
Figure G-2: Acquiring Attributes and Values.
Different orders of the goodness of the values of the attributes are available: it is increasing, when
the first value is the worst; decreasing, if the first is the best one; if one value is not better then the
other one, the order is nominal.
Tip: If you use the same value ordering for all attributes, it will make easier to define the
rules.
Once the attributes and their values are defined, if we are building a rule-based knowledge base, the
next step is to determine the dependencies between the attributes. This consists of two parts: the
which(s) and the how(s) of dependencies.
2.3 Hierarchy of the Attributes: the Rule-Based Graph
The which(s) attribute dependencies means to allocate for each attribute on which other attributes
it depends on. This is done by constructing a hierarchy of attributes called Rule-Based (or deductive)
Graph on the third pane of Doctus, named Rule-Based Graph. (See Figure G-3) To construct the
graph, drag-and-drop is used.
Figure G-3: The Rule-Based Graph.
Tip: Do not connect more then 3-4 attributes onto one node, to make yourself the rule input
easy.
If attribute B is connected onto attribute A (which means that A depends on B), then B is called
factor of A. The same attribute may factor of different attributes, though not to itself (directly on
G-10
indirectly). The root of the graph is not a factor of any other attribute; it is called decision attribute, or
outcome. The leaves of the graphs have no factors, they are the input attributes. There will be
attributes, whichs are factors of other attributes and having factors themselves; these are the dependent
attributes.
When the Rule-Based Graph is constructed, rules are to be defined in each node of the graph.
Tip: Once your Rule-Based Graph is constructed, dont start with defining rules, first acquire
cases. While acquiring cases the attributes and their values are likely to be modified.
2.4 Acquisition of Cases
Knowledge-based systems are used to reason about cases. Cases can be anything that we can
describe from all important aspects (i.e. defined attributes). One value of every attribute is assigned to
each of the cases. Actually one value is the default but Doctus can also handle Unknown, Dont
care and distributed values.
The acquisition of cases happens on the second pane of Doctus, named Cases. In deduction or
Rule-Based Reasoning it follows the construction of the Rule-Based Graph; however, new cases may be
added to the knowledge base at any time. To assign a value of an attribute to a case, use the context
menu from the right mouse-button. (See Figure G-4)
Tip: To simplify the acquisition of cases adjust the view to show the input attributes only.
Figure G-4: Acquisition of Cases.
2.5 Knowledge Import Data Mining
Knowledge Import is a feature to get cases directly from external data sources. Built-in types of
input sources are (some of them available in advanced mode only): Excel Workbook, Microsoft Query,
Mailbox and URL Encoded Cases, though any source may be accessed through ODBC. (See Figure G-
5)
G-11
Figure G-5: Knowledge Import.
Records from the data sources are retrieved as cases of the knowledge base. The attribute linking
and the case import are facilitated by a wizard. Apart from text type values, also called flexible values,
numeric input may also be used, handled by a clustering algorithm.
Tip: Some special characters in the field names of the data sources may not be recognized and
they mustnt contain spaces, thus pay attention to avoid them. Often a space at the end of the
field name remains unnoticed!
The conception of data mining evolved from the observation that organizations store a huge
amount of data (in the databases and data warehouses of their information systems) and use most of
them for nothing. It is presumed that new knowledge could be discovered finding the rules hidden in
relations between these data. The numeric data from the sources are to be transformed into symbols.
Once Doctus is connected to external data source via its Knowledge Import module, the
transformation is done using a built-in clustering algorithm. (See Figure G-6) The user defines the
number of expected clusters, the rest of the process is automated.
G-12
Figure G-6: Clustering External Data.
In the Rule-Based Reasoning the data mining concept is applied starting from the point that some of
the soft information in the head of the expert may be traced back to some soft relations between hard
data stored in databases or data warehouses. Therefore some of the branches of the Rule-Based Graph
may at the end (on leaves) have numeric input, namely symbols coming from clusters of a numeric
range. As the databases and data warehouses are updated constantly, the knowledge base is always up-
to-date in regard to data. Constructing a smart knowledge base an intelligent alerting function can be
created, which will call attention only when the changes of data change the output as well.
2.6 The Rules
Selecting an attribute on Attributes, Cases or Rule-Based Graph pane, the name of the fourth
pane changes, incorporating the name of the selected attribute in Rules of. (See Figure G-3) In
each node of the graph (so for each dependent attribute) a set of rules is to be given, to assign an
outcome (a value of the selected attribute) to each variation of the values of the factors.
If a rule connects one value of each factor, it is called elementary rule.
Maths: Use the markings: Attributes: A, B, C, (X is the decision attribute)
Values: A={a1, a2, a3, }; B={b1, b2, b3, }; X={x1, x2, x3, }
Rules: A=a1 B=b2 C=c1 X=x1
Read: If A is a1 and B is b2 and C is c1 and then X is x1
G-13
If a rule covers a range greater then one single value for at least one attribute, it is called complex
rule. The covered range may contain neighbour values only; it may be closed (between two values) or
opened (worst or better then one value).
Maths: Use the markings from above.
Rule: A[a2, a5] B=b2 C=c1 X=x1
Read: If A is between a1 and a5 and B is b2 and C is c1 and then X is x1
Rule: A a2 B=b2 C=c1 X=x1
Read: If A is better or equal to a2 and B is b2 and C is c1 and then X is x1
The complex rules can be seen as aggregations of elementary rules. The knowledge is easier to
describe if it is done by fewer complex rules. Of course the same knowledge can be described by
different sets of complex rules, i.e. the elementary rules can be variously aggregated.
Doctus provides two different surfaces to handle rules; the user can switch between them. On 1D
surface rules are presented in form of rule list, new rules may be defined editing them directly into the
table or using the insert new rule command. (See Figure G-7) On 2D surface some of the factors will
indicate the rows and others the columns of the table. Each cell of the table is a rule, its inputs are
defined by its position (row and column) and the user defines the output selecting a value from the
right-mouse-click-menu. (See Figure G-8) More than one cell can be selected at the same time.
Figure G-7: Rules in 1D.
G-14
Figure G-8: Rules in 2D.
Advanced: Input of rules in 1D may be also done by putting in a general rule (any of
factors may have any of its values) first and then splitting or dividing the ranges of the factors.
In 2D rules may also be defined by levels, which means to use logic if and then the
output is at least.
Doctus provides advices for the rule input using the previously defined order of values. (See Figure
G-9) Accepting the advices helps to avoid inconsistency of the rule set.
Figure G-9: Advices for Input of Rules.
Either using the advices or not nothing prevents from putting in an inconsistent rule. However, if
the option is selected, Doctus can flag the inconsistent rules. (See Figure G-10)
G-15
Tip: Under certain circumstances the inconsistent rules may appear to be logically correct. If it
happens to you consider changing the order of values.
Figure G-10: Checking the Consistency of the Rule Set.
Tip: Define the rules in 2D, using the advices and the consistency check, then switch to 1D
to discover new knowledge from complex rules.
2.7 The Reasoning
Reasoning in a Rule-Based system is done by triggering the rules for the cases, getting a value of the
decision attribute for each case; therefore it is also called evaluation of cases. The results may be seen
on the Cases and on the Rule-Based Graph panes. (See Figure G-11)
Figure G-11: Results of Rule-Based Reasoning.
G-16
Technical: In Doctus rules are stored in form of rule list, like in 1D view. While
reasoning, Doctus searches top-down the rule list for the appropriate rule; stopping if the first
one is found. It means that multiple coverage of a domain does not cause a trouble though if
there is no appropriate rule result remains Unknown.
Technical: The real computing is some more altered by Unknown, Dont care and
distributed values. If there is an Unknown case feature or rule outcome the evaluation of the
case will be Unknown as well. If there is a distributed value, the outcome of the rule set will
be distributed in the same proportion, if there are more distributed input values, the outcome
will follow the superposition of these distributions. A Dont care is considered to be an
equal distribution of all input values.
2.8 Decision Analyses and Fine-Tuning
To take a decision it would be nice to have one and only one case having the best value of the
decision attribute. However, it happens extremely rarely at first try, usually there are none or several
cases with the best output. To facilitate the fine-tuning of the knowledge base Doctus is equipped with
an explanatory option. (See Figure G-12).Explanation can be asked about the evaluation of every case
and for each of its dependent attribute as well; it shows which rule was used to get the outcome.
Figure G-12: Explaining the Results.
The explanation helps us to locate the reason of the unsatisfactory evaluation, though the
refinement is to be done manually. Apart from above mentioned another reason to fine-tune the
knowledge base can be to simplify the description of knowledge described by the knowledge base of
course without loosing anything important.
Fine-tuning usually implies one or more of the followings:

input of new attributes and/or values
renaming existing attributes and/or values
G-17
deleting existing attributes and/or values

modification of rules
modification of cases.
After refinement the knowledge base should reflect the opinion of the expert about the decision.
The fine-tuning is finished only when the expert agrees with all the elements of the knowledge base.
2.9 Knowledge Export Intelligent Portal
Doctus is capable of exporting knowledge bases in various forms of intelligent agents. (See Figure
G-13) Some of these features are available in advanced mode only.
Figure G-13: Knowledge Export.
The exported knowledge may be:

Static exported knowledge is some sort of snapshot of the knowledge base with no
reasoning capability. Different types of graphics and tables can be generated, presenting
graphs or details about the knowledge base. The report in rtf format, containing all tables is
also available.
Dynamic exported knowledge has capability of reasoning. Built-in templates are available to
generate webserver (php, perl, jsp), client (html pages with enabled on-page reasoning,
javascript) and desktop (c++, xls) applications.
Advanced: Build your specialized export templates based on the above listed predefined
ones.
Using the Knowledge Export feature the exported knowledge base can be made available to various
users, who will be able to use it for evaluation, though they will not be able to modify it. Some types of
the exported knowledge are also appropriate to be placed into portals in forms of portlets. If the
knowledge base was previously connected to external sources to retrieve data, this connection may be
G-18
maintained for the exported knowledge base as well. Thus quick evaluations can be made, when the
decision maker needs to fill in only a few fields (assign a value of an attribute for a case), while the rest
is retrieved from databases. (Figure G-14)
Figure G-14: Evaluation in Intelligent Portal.
3 Routine Decision (when there are few dozen cases in the domain)
When the expert cannot or do not want to define the rules, though he has a few dozen of cases with
evaluations in his experience, these cases can be used to find the relations between the values of the
attributes. The basis of the reasoning are the cases, therefore this kind of reasoning is called Case-Based
G-19
Reasoning. As the set of particular cases is used to find generalized rules, which describe it
appropriately, it is also called induction.
3.1 Benefits
The more obvious benefit of the case-based knowledge base is that the number of used attributes is
reduced, to the informative ones. This makes the deputation of a decision much easier. Naturally, it is
nothing of the sort of making programmed decision makers, as the Case-Based Graph represents the
only the experts experience at given conditions. If a new case appear, which cannot be described with
the knowledge base, it means there were no similar cases in the experts experience. The conditions
may also change. Thus it is highly recommended to add the new cases constantly to the knowledge
base, to maintain it as fresh as possible.
The greatest benefit of the building a case-based knowledge base is less obvious. This process is
almost always accompanied with knowledge discovery, that is to say it makes a part of tacit knowledge
explicit. It is very common that the expert is astonished at the first sight of the Case-Based Graph, thus
the fine-tuning is not only necessary to make subtle adjustments to the knowledge base but also to get a
deeper understanding of the result.
3.2 Acquisition of Attributes
Knowledge Acquisition always starts with formulation of the aspects of the decision. Aspects are
given by the expert as attributes (i.e. the names of the attributes) and their values. A value of an
attribute is a decision criterion.
Tip: Use short expressions of the special language of your domain of expertise; for better
understanding descriptions may be attached to each of the attributes and values.
The acquisition of attributes and their values happens on the first pane of Doctus named
Attributes. (See Figure G-15)
Figure G-15: Acquiring Attributes and Values.
Different orders of the goodness of the values of the attributes are available: it is increasing, when
the first value is the worst; decreasing, if the first is the best one; if one value is not better then the
other one, the order is nominal.
G-20
Tip: If you use the same value ordering for all attributes, it will make easier to define the
rules.
Once the attributes and their values are defined, if we are building a case-based knowledge base, the
next step is to acquire the cases, including the outcome for each case. Doctus will discover the rules,
which describe the cases of the experts experience.
3.3 Acquisition of Cases
Knowledge-based systems are used to reason about cases. Cases can be anything that we can
describe from all important aspects (i.e. defined attributes). One value of every attribute is assigned to
each of the cases. Actually one value is the default but Doctus can also handle Unknown, Dont
care and distributed values.
The acquisition of cases happens on the second pane of Doctus, named Cases. In induction or
Case-Based Reasoning it comes immediately after Acquisition of Attributes, however, new cases may
be added to the knowledge base at any time. To assign a value of an attribute to a case, use context
menu from the right mouse-button. (See Figure G-16)
Tip: To simplify the acquisition of cases adjust the view to show the lonely attributes only.
Figure G-16: Acquisition of Cases.
3.4 Knowledge Import Data Mining
Knowledge Import is a feature to get cases directly from external databases. Built-in types of input
sources are (some of them available in advanced mode only): Excel Workbook, Microsoft Query,
Mailbox and URL Encoded Cases, though any source may be accessed through ODBC. (See Figure G-
17)
G-21
Figure G-17: Knowledge Import.
Records from the data sources are retrieved as cases of the knowledge base. The attribute linking
and the case import are facilitated by a wizard. Apart from text type values, also called flexible values,
numeric input may also be used, handled by a clustering algorithm.
Tip: Some special characters in the field names of the data sources may not be recognized and
they mustnt contain spaces, thus pay attention to avoid them. Often a space at the end of the
field name remains unnoticed!
The conception of data mining evolved from the observation that organizations store a huge
amount of data (in the databases and data warehouses of their information systems) and use most of
them for nothing. It is presumed that new knowledge could be discovered finding the rules hidden in
relations between these data. The numeric data from the sources are to be transformed into symbols.
Once Doctus is connected to external data source via its Knowledge Import module, the
transformation is done using a built-in clustering algorithm. (See Figure G-18) The user defines the
number of expected clusters, the rest of the process is automated.
G-22
Figure G-18: Clustering External Data.
In the Case-Based Reasoning the data mining concept is applied based on presumption that the rule
discovery algorithm of Doctus can be also used to discover patterns between the data stored in
databases or data warehouses. The numeric data is first transformed into symbols using the clustering
algorithms, while non-numeric data are retrieved as flexible values of appropriate attributes. The result
of this kind of data mining is very easy to interpret for the expert of the domain of course.
3.5 Decision Tree: the Case-Based Graph
Doctus generates the Case-Based Graph classifying the cases acquired from the expert. The Case-
Based Graph is a decision tree; it does not show dependencies but the if then rules induced by
processing the cases. The if then rules may be read from the root of the graph towards its leaves,
where the value of the outcome is shown. (See Figure G-19)
There are three alternative branching methods to generate the Case-Based Graph: The default is
called Efficient, which is described in the following chapter. The Bipolar makes two branches for
each node, grouping the values of the attributes to bad and good. The Heuristic provides the same
result as Efficient if there is a great amount of cases and/or attributes, which would otherwise highly
increase the computing time.
G-23
Figure G-19: The Case-Based Graph.
The attributes appearing in the Case-Based Graph are called informative attributes, as they are
sufficient to classify all the cases.
3.6 Classification of Cases
After all, how the Case-Based Graph is constructed? Lets presume that all cases form a disordered
set, where the order is defined as homogeneity by benchmark values (values of outcome attribute),
which means that cases in one subset have the same benchmark value. The attribute is searched, which
contributes the most to the order. The attributes are taken one-by-one forming subsets according to
their values. Their strength in making order is measured by an entropy-gain (informativity) calculating
algorithm. The most informative attribute is chosen (the root of the graph) and the first level subsets
are formed according to its values. These subsets are further divided using the same algorithm until all
subsets are homogenous by benchmark values. When a homogenous subset is formed, it is not further
divided; it will be a leaf of the graph.
G-24
Maths: Determining informativity can be found in Appendix Determining

Informativity (R-7.4).
Advanced: The described algorithm could lead to infinite cycles; therefore thresholds are
defined to terminate the process in these situations. The thresholds can also be used to control
the size of the Case-Based Graph. The Minimal Informativity Density is in direct ratio to
the informativity and to the weight of cases in the node. The Maximal Majority of a
Benchmark Value is the upper limit for the proportion of the dominant benchmark value in
the node. The Minimal Weight of Cases in Node indicates the minimal ratio of the cases in
a node against the starting number of cases needed to branch the node.
Technical: Real computing is some more altered by Unknown, Dont care and
distributed values. A case with Unknown value of an attribute is neglected when that
attribute appear on the Case-Based Graph (generated or chosen). In the same situation a case
with a distributed value falls into the case subset of each of the values keeping the proportion of
the distribution. Cases containing Dont care are considered to have all the input values.
3.7 The Reasoning
The result of the Case-Based Reasoning is the Case-Based Graph, which describes the rules induced
from the cases of the experts experience. It is easy to reason about new cases using the Case-Based
Graph as well: the new case simply has to be positioned according to its features by the informative
attributes, following a path from the root to a leaf of the graph. However, classification of new cases in
Doctus is facilitated with reduction (G-4) and with some of the Knowledge Export (R-4.1) solutions.
3.8 Decision Analyses and Fine-Tuning
It is usually not easy for the expert, that his experience may be described with only a few of the
attributes he defined. Analysis of the Case-Based Graph is facilitated with hands-on information
provided by Doctus about the informativity, density, cases and statistics for the nodes of the graph.
(See Figure G-20) It is easy to change the attributes in the nodes of the graph, though there are
conditions, which are likely to be observant of. The fine-tuning is switching between the parallel or
nearly parallel knowledge models. This means that the swaps of the attributes in the nodes of the graph
are justifiable only if they are of equal or nearly equal informativity and density.
G-25
Figure G-20: Informativity and Density of the Attributes.
There are also another ways of fine-tuning: Sometimes a case is found, that cannot fit the set and
makes serious degenerations to the Case-Based Graph. These cases usually cannot be described with
the attributes defined, thus we call them odd-one-outs. The solution for these is to be excluded from
the set used as bases for Case-Based Reasoning. Sometimes two (or more) cases are found, that are
completely the same, except for the benchmark. It usually means that a new attribute or a new value is
needed to be defined, which distinguishes the cases in question. The cases themselves may be modified
as well.
Tip: For better understanding it can be very useful to generate Case-Based Graphs with
different benchmarks. This highlights the interdependencies of the attributes for the set of cases.
Using the Bipolar branching method shows which values of the attributes are more similar,
and are likely to indicate possibility to reduce the number of values.
G-26

also available.
ones.
the exported knowledge are also appropriate to be placed into portals in forms of portlets. There are
two very helpful usages of exported case-based knowledge bases: they can be used to collect new cases
to extend the case base (Figure G-22 left) and to give a quick evaluation following a path from the root
of the Case-Based Graph to its leaf. (Figure G-22 right)
G-27
Figure G-22: Exported Case-Based Knowledge Base used for case collection and evaluation.
4 Learning from Cases (reduction of the model)
Once the expert accepted the Case-Based Graph, a new rule-based knowledge base can be created,
which contains only the informative attributes but gives the same evaluation for the cases as the ones
used for the induction of rules. The reasoning uses rules but they are induced from the set of cases,
thus this type of reasoning is called Case-Based Rule Reasoning. As the knowledge base is generated
automatically by reducing an existing model, it is also called reduction. The attributes and the cases are
already acquired, therefore there are no chapters discussing acquisition of attributes, acquisition of cases or
knowledge import.
4.1 Benefits
The great benefit of the Case-Based Rule Reasoning is the reduced size, i.e. the significantly
decreased number of the attributes. It enables the user to make a quick evaluation of new cases but
attention is to be paid to possible loss of actuality. To avoid the use of outdated knowledge base, the
original case-based knowledge base is to be maintained, constantly adding the new cases and
regenerating the Case-Based Graph. If the conditions are changed, the Case-Based Graph will alter.
G-28
Tip: If having a rule-based knowledge base and, since its been in use, a number of cases are
accumulated, perform the Case-Based Reasoning on the same knowledge base and then extract
the rules generating a reduced knowledge base. Doing so, you are able to densify the knowledge
described by the primer rule-based knowledge base.
4.2 Single-Level Hierarchy
The rule-based knowledge base generated by Doctus from the Case-Based Graph forms a single-
level hierarchy of attributes. To distinguish this special kind of Rule-Based Graph from the ordinary
ones, a different name is dedicated to it, which also appears on the third pane of Doctus (Figure G-23):
Case-Based Rule Graph.
Apart from being single-levelled, this Rule-Based Graph can be used similarly to the ones built in
ordinary rule-based knowledge bases, see chapter Original Decision Hierarchy of the Attributes: the
Rule-Based Graph (G-2.3).
Figure G-23: The Case-Based Rule Graph.
4.3 Missing or Indefinite Rules
If there were value ranges of some rules not covered or multiply covered by cases used for Case-
Based Reasoning, in the rule set of the reduced knowledge base some rules may be missing or
indefinite. (See Figure G-24)
G-29
Figure G-24: Indefinite Rules in Reduction.
The missing or indefinite rules may indicate impossible range or not well-defined attributes or
values. Usually fine-tuning is needed to make these situations clear. The available operations of the rule
set are the same then in rule-based knowledge bases, see chapter Original Decision The Rules (G-
2.6).
4.4 The Reasoning
Reasoning in case-based rule system works and looks the same as in rule-based systems (see chapter
Original Decision The Reasoning (G-2.7)), though without fine-tuning the evaluation of a new case(s)
may be indefinite or none at all. In this second situation it is strongly recommended to repeat the Case-
Based Reasoning with the new case(s) included.
4.5 Tacit Knowledge and Fine-Tuning
The missing or indefinite rules may be made definite by simply changing the outcomes of the rules
manually. However, it is worth consideration, what caused these missing or indefinite rules? If the
expert is sure, that it indicates an impossible range, the rule may remain missing or indefinite; if there is
a new case(s) falling into that range, perhaps the conditions of the reasoning are changed, thus the
refreshment of the Case-Based Reasoning should be considered.
If during the fine-tuning of the reduced knowledge base implied changes of attributes and/or values,
these changes should be applied to the case-based knowledge base as well, and the Case-Based
Reasoning should be repeated.
As the hierarchy of attributes in the reduced knowledge base is single-levelled, it can easily happen
that there are more then 3-4 attributes, which makes handling of the rule set difficult. There is no
obstacle to modify the graph into a multi-level one, using the same drag-and-drop technique as
described in chapter Original Decision Hierarchy of the Attributes: the Rule-Based Graph (G-2.3).
Tip: Before modifying the Case-Based Rule Graph, repeat the Case-Based reasoning choosing
different benchmarks, for deeper understanding of the interdependencies of the attributes.
Fine tuning the Case-Based Rule Graph and using it as feedback to the original rule-based or case-
based knowledge base the tacit knowledge is pulled to explicit domain.
G-30

also available.
ones.
the exported knowledge are also appropriate to be placed into portals in forms of portlets. As reduction
produces a rule-based knowledge base the exported versions of it are very similar to the deduction, the
only difference is that here we usually have less fields to fill. (Figure G-26)
G-31
Figure G-26: The exported reduced model.
G-32
Doctus Reference http://www.doctus.info
Doctus Reference
1 A Knowledge-Based System
Doctus is a Knowledge-Based System. It inputs, stores, examines, manages and transforms

knowledge from arbitrary domains of life. In turn the knowledge must be represented in a special
symbolic form for Doctus.
All of the aspects are to be set out with their categories indicated. The objects of the knowledge
domain (i.e. the decision alternatives) are described by these categories. After that we need only
symbolic logic, which is put into action by a piece of software.
2 The Five Panes of Doctus
The knowledge bases opened in Doctus are presented in a child windows of the main frame
(Multiple Document Interface MDI). Each window has five panes to display different views of the
same knowledge base:
All of these panes are described according to the following scheme:

View What can be seen on this pane?
Edit What can be entered or modified on this pane?
Examine What can be searched for or checked on this pane?
Manage What can be generated on this pane?
2.1 Attributes
The aspects of the knowledge domain, which are called attributes in Doctus, are listed on the
Attributes pane. Each attribute consist of its name and its values to define the categories of the
aspect. The values are symbols (i.e. words or expressions), for numeric aspects of domains a clustering
algorithm is provided to transform them into symbols.
R-33
Figure R-1: Competitors can be classified based on the amount of their done works into four categories. They are none,
some, mean and many.
Each attribute stretches out a dimension of the knowledge domain. Although it is not continuous,
but discrete by its values, proper number of attributes allows to describe even complex domains. The
clear human meaning of the attributes/values maintains order among the interpretation.
2.1.1 View
There are a couple of other properties of each attribute, which can be shown on this pane.
2.1.1.1 Description
Each attribute and each value can have an own description eventually with many lines.
Choose Description from the menu View. A special resizable edit box appears to show and edit
the description of the attribute or value currently selected on Attributes pane. Ctrl-Up and Ctrl-
Down allow stepping it to the previous and the next one respectively without the need of leaving the
box.
Tip: Keep names to be short to make convenient using other functions of Doctus. Put broader
explanations into the description.
Advanced: Description can hold special properties too, as they can be interpreted by a
corresponding knowledge export template (R-4.1). The template syntax offers numerous string
manipulation operations to extract them from the description with a suitable layout.
2.1.1.2 Customize
Choose Customize from the menu View.
R-34
Figure R-2: Customize Attributes pane
This dialog controls the appearance of the Attributes pane.

Vertical Attributes are presented as rows by default. If it is checked, they appear as columns.
Wrap Value properties are listed in one row, in the row of the attribute. Should this layout be
too wide, Wrap directs them into one column.
Attribute Type Flags The small icons showing the attribute type on the right of the attribute
names can be enabled or disabled by this switch.
AutoFit Text Cell widths are adjusted automatically to the width of the cell text if this
checkbox is on.
Attribute & Value Properties The appearance of attribute and value properties on the
Attributes pane can be enabled/disabled by selecting them from lists. The meaning of each
of them is described in Attribute Properties (R-2.1.2.8).
Tip: If more value properties are shown, use Wrap to enhance readability.
2.1.2 Edit
2.1.2.1 Create Attributes & Values
Move cursor to the left head and type the name of the attribute. Leave cell with Tab or Enter,
and input the name of the first value. Repeat this until all the values are entered. Ctrl-Enter moves the
cursor to the start of the next row. Off course mouse buttons may be also used in usual way.
Advanced: If a new value is added to a factor attribute, value ranges are unaffected, unless
it is a full range, which is expanded with the new value to remain full range.
Technical: The performance of some knowledge processing functions is decreased

dramatically after about dozen values. The used algorithm falls prey to a combinatorial
explosion. Anyway the attributes needing so many values are really rare.
2.1.2.2 Rename Attributes & Values
Type the new name on the cell and leave it. The attribute or the value is renamed, additionally all
occurrences are updated automatically.
R-35
2.1.2.3 Insert Attribute
Move cursor to the left head or select the entire row and choose Insert Attribute from the menu
Edit or from the context menu. A new attribute is inserted before the current one with a default
name. Type the new name to overwrite it and leave the cell.
2.1.2.4 Insert Value
Move cursor to a white cell and choose Insert Value from the menu Edit or from the context
menu. A new value is inserted before the current one with a default name. Type the new name to
overwrite it and leave the cell.
Advanced: If a new value is inserted into an attribute that is a factor of another one, the
value range of it is expanded by this new value. Therefore a new, uncovered domain appears in
the rule set, unless the range is a full range or the range would be splitted by the new value. In
these cases the rule will cover the expanded domain.
2.1.2.5 Delete Attribute
Select the rows of attributes to delete and choose Delete Attribute from the menu Edit or from
the context menu. The selected attributes and all corresponding knowledge are deleted.
Advanced: The corresponding knowledge consists of values, case features and rules of this
attribute. The rule-based attribute dependencies are broken.
Advanced: If an attribute, which is a factor of another one is deleted, the rule domain of it
loses a dimension, but no rules are deleted. So the rule set becomes less distinctive, thus
multiple covered rule domains are likely to arise.
2.1.2.6 Delete Value
To delete values move cursor to or select multiple cells of values and choose Delete Values from
the menu Edit or from the context menu. The selected values and all corresponding knowledge are
deleted.
Advanced: The corresponding knowledge consists of case features and rule outputs with the
values to be deleted. First, Doctus searches the knowledge base, whether they are used, and the
delete is confirmed with a warning message. These values are replaced by Unknown.
Advanced: If the attribute of which the value is deleted, is a factor of another attribute, the
value range of it is reduced, but no rules are deleted. Actually some rules can loose their value
ranges at all, in other words they have empty range, so there is no domain anymore, where they
can be active. Accordingly, these rules can not be seen on 2D view (R-2.4.1.3) of the rule set,
so they can be deleted on the rule list view (R-2.4.1.2) or by Rules Purify Rule Set (R-
2.4.2.11).
2.1.2.7 Move Attribute
Select rows of attributes to move, grip the selection at its border, drag and drop it. The attributes are
reordered. For other selections the normal text mode move (R-3.6) is active. Currently there is no way
to reorder values.
R-36
2.1.2.8 Attribute Properties
Move cursor to or select multiple attributes, then choose Attribute Properties from the menu
Edit or from the context menu. A multi-tab dialog box openes to show and edit attribute properties.
2.1.2.8.1 Value Ordering
Figure R-3: Attribute properties Value Ordering
Most attributes make sense of some goodness. The values of such attributes can be set out in the
order of this goodness. If it is done so, select Increasing for increasing goodness. Now the leftmost
value is the worst one, and the rightmost is the best one. Select Decreasing for the opposite order.
Select Nominal, if attribute has no sense of goodness, so the values are only distinguished.
Tip: Use Increasing to have a clearly arranged rule domain if appropriate, it will make the
rule input easier.
Advanced: Selecting Nominal for an attribute disables consistency (R-2.4.1.4) check for
that attribute only. Selecting Nominal for an attribute that is a factor of another one,
disables consistency check for this attribute concerning that factor only.
2.1.2.8.2 Branching Method
Figure R-4: Attribute properties Branching Method
R-37
These properties control the generating behaviour of the Case-Based Graph. The general branching
method selection (R-2.5.4.1.1) can be filtered by these properties attribute-specifically. In the left
column the allowed branching method can be selected:
No Branching None of the branching methods are enabled. This attribute is left out from the
possible informative attributes.
Polar Only Always the Polar method is used for this attribute.
Disable Efficient Heuristic is used instead of Efficient, the Polar is used regularly.
Any All methods are enabled, so the general branching method selection is used for this
attribute.
Tip: Choose Disable Efficient for attributes with many values to increase performance of
graph generation.
The column on the right fixes the general ordering selection (R-2.5.4.1.1) for this attribute:
Default Use the general ordering selection (R-2.5.4.1.1).
Normal Do not use Ordered for this attribute.
Ordered Use Ordered for this attribute.
Tip: Choose Ordered for range of values such as less than 10, between 10 and 20 and
more than 20, if you want to avoid leaky value selections, such as less than 10 or more than
20.
2.1.2.8.3 Automatic Values
Figure R-5: Attribute properties Auto Values
Knowledge Import (R-4.2) of Doctus is featured with automatic value gathering:

Constant Values No gathering. Only the predefined values are allowed. The undefined value
symbols are replaced by Unknown.
Flexible Values Simple value gathering. All undefined values (incliding the numbers!) are
added as new values of the attribute.
R-38
Numeric Cluster Analysis Values are treated as a decimal floating point numbers, submitted to
numeric cluster analysis, and replaced by the nearest cluster value. Press Configure to
configure cluster analysis. For details see Cluster Configuration (R-2.1.2.8.4).
2.1.2.8.4 Cluster Configuration
Enable Numeric Cluster Analysis (R-2.1.2.8.3) for the attribute, then press Configure in the dialog
or choose Cluster Configuration from the menu Edit.
Figure R-6: Attribute properties Cluster Configuration
This dialog shows the settings for cluster analysis and the cluster state. The cluster analysis is
processed by starting Knowledge Import (R-4.2); thus the cluster state shown in the table of the dialog
box is generated automatically, however it can be modified manually. The cluster analysis can be set up
by the Link Wizard of Knowledge Import (R-4.2.6) too.
The settings for cluster analysis are placed in the dialog below the table:
Number of Values Expected Cluster analysis results this number of clusters for this attribute.
Each cluster represents a value.
Error measure The method to compute error distance for clustering and choosing the best
cluster pair to link during cluster analysis. Currently only Euclidean supported, which is the
sum of squares of differences of standardized data among all dimensions.
Amalgamation/linkage rules The way to reduce clusters to the expected number. Currently only
K-Means is supported, which searches exhaustive the nearest two clusters to link.
Insert Dimension A new column is created in the table for a new dimension. Each dimension
can have a different external source, so each of it has independent cluster state. The external
source for this dimension can be select in the column head.
Remove Dimension The column of the current dimension is removed. The last dimension can
not be deleted.
R-39
Technical: Knowledge Import reads the first 1000 cases as initial clusters into a buffer,
and starts to reduce the number of clusters by the chosen amalgamation/linkage rule
repeatedly until the required number of clusters is reached.
The first two rows of the table show the current state of clusters:
Total Cases The total number of cases analysed.
Total Error The total error encountered by clustering all the cases.
Average The average of data of all clustered cases for this dimension. It is used for
standardizing.
Standard Deviation The standard deviation of the data of all clustered cases for this
dimension. It is used for standardizing.
Starting with the third white row the clusters are displayed. Each row of a cluster has the following
cells.
Left head cell The name of the value representing this cluster, can be overwritten manually.
Cases The number of cases taken into this cluster.
Average Error The average error encountered by taking cases into this cluster.
Value The center of this cluster. This is averaged from the data of cases taken into this
cluster for this dimension.
2.1.2.8.5 Knowledge import
Figure R-7: Attribute properties External Input
Knowledge Import (R-4.2) of Doctus fills case features for the attributes linked to an external
source. This tab of the dialog box shows and edits this linkage, although it is normally set by the Link
Wizard of Knowledge Import (R-4.2.6). The interpretation of these values depends on the data source
of Knowledge Import (R-4.2).
Table The table name (Microsoft Query & ODBC) or the sheet name (Excel workbook)
from which the external data is read.
R-40
Column The name of the column of the table from which the external data is read for this
attribute. If more than one dimension is set for clustering, it has to be set on the column head
of the Cluster Configuration dialog (R-2.1.2.8.4).
Case Name The name of the column of the table from which the case identifier of the
external data is read. Each attribute that is read from the same table, must be configured to the
same case name. If no Case Name is configured, external data is read in its natural order into
the Cases pane.
2.1.2.9 Set to Decision Attribute
A knowledge base actually can have more than one decision attributes but one of them is assigned as
the Decision Attribute. This attribute represents the conclusion of the deduction and the root node
of the Rule-Based Graph that is shown on its top. The row (or the column if vertical view is chosen) of
the Decision Attribute is highlighted with head background on the Attributes and the Cases
panes.
If the attribute, which is a factor of another one, is set to decision attribute, first it is pruned.
2.1.3 Examine
2.1.3.1 Search Unused Value
Choose Unused Value from the menu Search. All values that are neither used as case features
nor as rule outputs, are selected. It is safe to delete (R-2.1.2.6) all such values, but rule domains can be
damaged.
2.1.3.2 Search Factors
Choose Factors from the menu Search. All factor attributes of the current one are selected.
2.1.4 Manage
Not applicable.
2.2 Cases
The objects of the knowledge domain (i.e. the decision alternatives), which are called cases in
Doctus, are listed on the Cases pane. Each case has a name and a case feature for each attribute to
categorize the case according to the aspect represented by the attribute.
R-41
Figure R-8: The competitor 1 has some done works. This is vertical view.
A case feature can be:

normally a value of the corresponding attribute,
an empty cell representing Unknown,
a star (*) representing Dont care,
or a splitted cell representing a feature that is distributed by the probabilities of more values.
2.2.1 View
2.2.1.1 Description
Choose Description from the menu View. The description (R-2.1.1.1) of the attribute or the
value selected by the cursor position is displayed.
2.2.1.2 Customize
Figure R-9: Customize Cases pane
This dialog controls the appearance of the Cases pane. Following switches are supported:
R-42
Vertical Cases are presented as rows by default, built of the case features. If Vertical is
checked, cases appear as columns.
Wrap Rows of case features can be very wide if numerous attributes shown. Wrap wraps
these rows, so cases are divided into more rows, and more attributes take place in the same
column. Always the cases are wrapped, so the columns, if Vertical is on.
Attribute Type Flags The small icons showing the attribute type on the right of the attribute
names can be enabled or disabled by this switch.
checkbox is on.
In addition it can be controlled, which attributes are shown on this pane.

All All attributes.
Selected Attributes selected by the list box on the right. All attributes of the same type can be
selected by checking the appropriate checkbox on the left of the list box. If none of list box
items are selected, the mode All is activated.
Current The currently selected attribute.
Tip: If only the currently selected attribute is shown, press Ctrl-Up and Ctrl-Down to
step it without leaving this pane.
2.2.2 Edit
2.2.2.1 Create Cases
Move cursor to the left head and type the name of the case. Leave cell with Tab or Enter, and
enter the case features (R-2.2.2.2).
2.2.2.2 Enter case features
Move cursor to, or select cells to set to the same value, then follow actions in the next to set case
features to:
an exact value Choose a value from the context menu or press the key 1 for the first value,
2 for the second one, and so on.
Unknown Leave the cell blank, choose Delete Value from the Edit or context menu, or
press Del.
Dont Care Choose Set to Dont Care from the Edit or context menu, or press *.
a value distribution Choose Distribute from the Edit or context menu and set the rate (R-
2.2.2.3),
R-43
Tip: To collect values of the attributes during the case features input, set Automatic Values
(R-2.1.2.8.3) of the attributes to Flexible, and paste textual data from the clipboard. For
manual input choose Edit from the context menu and type the value name, even if it is a
new one. The undefined values will be added to the attribute.
Advanced: The case features for dependent attributes can be set only by reasoning, though
they can be cleared manually by choosing Delete Value from the menu Edit or from the
context menu. Purpose of this is to purge possibly invalid reasoned case features from the
knowledge base.
2.2.2.3 Enter distributed case feature
Choose Distribute from the Edit or context menu.
Figure R-10: This case feature is set to some and mean in the rate of 1/3 to 2/3.
Following columns are displayed:

Left head Attribute name is shown on the top and the left lists its values of which the
distribution rate can be set.
Count The rate numerator of the value. This value can be overwritten manually. The
denominator can be set below the table.
Rate The rate of the value in probability form, which is within 0...1. This value can be also
overwritten manually.
Bar indicator The graphical indicator of the rate. Select the cell, and use slider below to set it.
If Keep Together is checked, the rate of all values but the modified one are kept in their original
proportion so, that the sum of the rates remains. In other words while the rate of a value increases, the
other ones proportionally decreases and vice versa.
2.2.2.4 Rename Cases
Type the new name on the head cell and leave it. The case is renamed.
R-44
2.2.2.5 Insert Case
Choose Insert Case from the Edit or context menu. A new case is inserted before the current
one with a default name. Type the new name to overwrite it and leave the cell. Input its features (R-
2.2.2.1).
2.2.2.6 Delete Case
Select the rows (or columns if vertical is view is chosen) of cases to delete and choose Delete Case
from the Edit or context menu, or press Del. The selected cases are deleted.
2.2.2.7 Move Case
Select rows of cases (or columns if vertical is view is chosen) to move, grip the selection at its
border, drag and drop it. The cases are reordered.
2.2.2.8 Move Attribute
Select columns of attributes (or rows if vertical is view is chosen) to move, grip the selection at its
border, drag, and drop it. If all attributes are shown, the attributes are reordered, like it would be done
on the Attributes pane. If selected (R-2.2.1.2) attributes are shown, only the showing order is
changed.
For other selections the normal text mode move (R-3.6) is active.
2.2.2.9 Paste External Data
Copy cells of a spreadsheet application or of a text processor to the clipboard. Switch to Doctus.
Move cursor to the upper left cell of the Cases pane, and choose Paste External Data from the
menu Edit. External data is imported from the clipboard. Values are collected for flexible (R-
2.1.2.8.3) attributes, and numeric input is clustered for numeric (R-2.1.2.8.4) attributes.
Technical: The clipboard format that is internally used to copy cells is a Tab delimited
textual form. It is supported by a number of spreadsheet applications and text editors, so they
can be used to prepare external data to import.
Advanced: Clustered attributes with multiple dimensions need more than one column,
actually one for each dimension.
2.2.2.10 Sort
Select column of attributes to sort by in precedence order, and choose Sort from the menu Edit.
Cases are sorted by the selected attributes.
2.2.3 Examine
2.2.3.1 Search Value Combination
Choose Value Combination from the menu Search. This dialog is opened:
R-45
Figure R-11: Searching for the competitors of which Tender is fair at least.
Select table cells of values to search for in the dialog. Select more values of an attribute to search for
all of those values (relation or). Select none or all values of an attribute to omit the attribute from the
search criteria. Select values of more attributes to filter by more attributes (relation and). All cases with
the chosen values are selected.
Choose Factors from the menu Search. All factor attributes of the current one are selected.
2.2.3.3 Search Explanation
Move cursor to a case feature of a dependent attribute. Choose Explanation from the menu
Search. The Rules pane is activated and the rule is selected, which by the case feature is reasoned
(R-2.2.4.1).
2.2.4 Manage
2.2.4.1 Deductive Reasoning
Choose Deductive Reasoning from the menu Knowledge Management. All case features of all
dependent attributes are overwritten by values that are reasoned by the appropriate rules. If not all
attributes are shown, the decision attributes are enabled to show the resulted case features.
The reasoning can be limited by selecting certain cells. Only the selected cells are reasoned.
R-46
Advanced: It is not ensured that all case features are actual, which act as input of the used
rules, since they are not reasoned if they are not selected.
Advanced: The behaviour of special case features during the reasoning are detailed in Rules
Rule List & Rule Processing (R-2.4.1.2).
2.2.4.2 Accept Advice
The plausible advice for a rule set estimates the outputfor the uncovered domain based on other
rules and consistency considerations (R-2.4.1.4). Choose Accept Advice from the menu Knowledge
Management to accept this estimation by deductive reasoning.
Tip: If a rule set results Unknown even if Accept Advice is on, check the input case
features whether they are Unknown.
2.3 Rule-Based Graph
Attributes can depend on each other. These dependencies build a hierarchy, which is depicted by the
Rule-Based Graph.
Figure R-12: Tender depends on Reference and Finance, Finance depends on Warranty and Price.
The Rule-Based Reasoning produces new case features based on rules and the case features of the
input attributes.
2.3.1 View
Each node represents an attribute from the Attributes pane. The decision attribute stands on the
top of the graph. The factor attributes are connected from below to the nodes of their dependent
attributes. This edge represents that the attribute depends on its factors. Input attributes are the leafs of
the graph, they have no factors.
Even if more decision attributes exist, only one graph is shown, which is for the decision attribute
(R-2.1.2.9) representing the conclusion of the knowledge base. The other ones are placed without their
factor nodes below the graph.
R-47
2.3.1.1 Customize
Figure R-13: Customize Rule-Based Graph pane
This dialog controls the appearance of the Rule-Based Graph pane. Following switches are
supported:
Horizontal Graph is mirrored that the root node is on the left of the window, and the factors
are on the right of it.
Condensed Nodes with more than two leafs can take a wide area of the screen. Check this
checkbox to arrange such leafs into two rows to reduce total width of the graph (in horizontal
view it will arrange the leafs into two columns to reduce the hight).
Additional nodes Select it, whether all other attributes, only the other decision attributes, or no
other attributes to be sown below the graph.
Optionally some additional information can be placed in the nodes of the attributes. Following
Node Legends are supported:
None No node legends.
Values Values of the attribute are listed in the nodes.
Cases Selected cases features are listed in the nodes.
Cases with Names Selected cases names and features are listed in the nodes.
Use the list box on the right to select cases to show for the last two legend mode. Additionally the
appearance of node legends can be limited by the following ways:
For a Level Only Check this checkbox to show node legends only at the current node and at
the nodes on the same level. Select a node on another level and choose Show legends on this
level from the context menu to change the current level to show legends on.
For Input Attributes Only Check this flag to show node legends only on leaf nodes.
If both of these checkboxes are set, legends are shown only on leaf nodes with the current level
(and relation).
R-48
2.3.1.2 Other Decision Attributes
A knowledge base can have sometimes (actually without attribute dependencies always) more than
one decision attributes. To see the graphs of the decision attributes other than the current one, ensure
to display them as Additional nodes, than set it to Decision Attribute (R-2.1.2.9).
2.3.1.3 Hide Above
To see only a certain branch of a huge Rule-Based Graph select the node to show it as root and
choose Hide Above from the context menu. The parent and its factor nodes are hidden. Only a short
edge indicates that the node is not the real root.
To see the full graph again, choose Hide Above once more, or set the original root node shown
below the graph to the Decision Attribute (R-2.1.2.9).
2.3.1.4 Hide Below
To hide factor nodes of a node choose Hide Below from the context menu. The factor branches
are hidden only some short edges indicate that the node is not a leaf.
To see the factors again, choose Hide Below once more.
2.3.2 Edit
2.3.2.1 Connect
To connect a factor node to another attribute node, drag the node to connect from below the graph
to the node to connect to. A dynamic edge notifies what connection is about to establish by dropping
the node.
Advanced: The rule domain of the dependent attribute obtains an additional dimension by
connecting the new factor. Rule domain is expanded by inserting full ranges of the new factor,
thus the rule set remains independent of it.
2.3.2.2 Arrange
To change the order of the factors drag one of the factor nodes holding down the Shift key. The
factor will neither be pruned nor connected to another node. The order of factor nodes defines the
order of factors on the Rules pane.
2.3.2.3 Prune
To disconnect a factor node, first select it, then choose Prune from the Edit or context menu. It
is removed from its dependent node and appears below the graph.
R-49
Advanced: The rule domain of the dependent attribute loses a dimension, but no rules are
deleted. So the rule set becomes less distinctive, thus multiply covered domains will arise.
2.3.3 Examine
2.3.3.1 Explain
Set node legends (R-2.3.1.1) for Cases with Names and select a dependent node. Choose Explain
from the context menu and select a case to explain its reasoned feature from the submenu. The rule, by
which the case feature is resulted, is selected on the Rules pane.
2.3.4 Manage
2.3.4.1 Deductive Reasoning
Set node legends (R-2.3.1.1) for Cases with Names and choose Deductive Reasoning from the
menu Knowledge Management. The case features are overwritten by reasoned values.
Advanced: The behaviour of special case features during the reasoning are detailed in Rules
Rule List & Rule Processing (R-2.4.1.2).
2.3.4.2 Accept Advice
See Cases Accept Advice (R-2.2.4.2).
Cases Deductive Reasoning (R-2.2.4.1) and Cases Accept Advice (R-2.2.4.2) are applicable with
the difference that the result can be shown in the node legends.
2.4 Rules
Each dependent attribute has a rule set that assigns a rule output to each value combination of its
factor attributes. So each factor stretches a dimension of discrete values in the rule domain. The
volume of this domain is the total number of the value combinations (say, elementary rules). It is
multiplied from the size of the dimensions, which is the number of the factor values. Thus it can be
very high, therefore Doctus stores no elementary, but complex rules, which cover not certain points
but rectangular volumes of the rule domain.
Consequently each complex rule has a value range for each of its factor, and it has a rule output that
is similar to a case feature (R-2.2.2.2).
2.4.1 View
2.4.1.1 Customize
R-50
Figure R-14: Customize Rules pane
This dialog controls the appearance of the Rules pane. Following switches are supported:
2 Dimensional Switch with this checkbox between Rule List (R-2.4.1.2) and 2 Dimensional
(R-2.4.1.3) viewing mode. The cursor position and the selection are transformed logically.
checkbox is on.
Flag Inconsistent Rules Check this box on to make Doctus check newly added rules by
consistency (R-2.4.1.4) against the existing ones. Eventual conflicts are indicated by small faces
of different moods on the affected rules.
Show Plausible Advice On 2D view the uncovered rule domain is seen as empty cells. Check
this switch on to display the possible consistent (R-2.4.1.4) rule outputs in place of it. It is
drawn with grey to distinguish it from real rule outputs.
2.4.1.2 Rule List & Rule Processing
Rule list view shows the rules as they are stored. The rules are listed in rows. The rightmost column
displays the rule output, and the ranges for the factors are shown on the left of it.
Figure R-15: The Rule List view of a rule set with five rules.
R-51
Figure R-16: The illustration of the rule set above.
How is the rule set triggered by deductive reasoning (R-2.3.4.1)? The rules are listed in processing
order. The topmost rule is checked first, whether it matches the case features for the factor attributes.
If yes, the adequate rule is found, else the search is continued at the next rule.
Example: The first rule of Figure R-15 says: If Reference is excellent and Finance is
good at least, then Tender is excellent. If not, the second rule is taken.
Figure R-16 shows the same rule set by 2 dimensional colored pies. Both factors, so
Reference and Finance have five values, they are marked by marbles. The marble at far
back represents excellent Reference, while the one at near in front is for excellent
Finance. The purple pie represents the first rule. The blue one is the last rule covering the
whole domain. Considering the processing order, an upper rule hides the other ones below
them. More multiple covered domains are shown.
If no matching rule is found, so an uncovered domain is questioned, Unknown is resulted, unless

advising is enabled (R-2.2.4.2). The behaviour is the same, if the case feature for one factor is
Unknown.
Technical: If a case feature for a factor is Dont care, it is replaced by its values one by
one, a rule is got for each of them, and the outputs are averaged.
Technical: If a case feature for a factor is distributed, it is replaced by the possible values
one by one, a rule is got for each of them, and the outputs are summarized proportional to the
distribution.
Maths: Deductive reasoning uses the theorem of conditional probability (Bayes theorem).
2.4.1.3 2 Dimensional
The active domain of the rule set is shown in this view mode. The rule domain, which has as many
dimensions as factors, is projected into 2 dimensions, as flat layers side by side. If there are more then 2
factors, some of them will appear in rows, some in columns.
R-52
Figure R-17: The same rule set as above in 2D view.
Figure R-18: The illustration of the above rule set. It is the same, isnt it?
Each elementary rule has an own cell for its output. All points of the domain have a cell, even it is
not covered by any rule. Hidden domains are not showed on this view.
Example: The pink rule of Figure R-18 has a small hidden domain under the purple one.
So its rule domain is larger than its active domain, which is shown selected on the Figure R-
17.
The dimensions can be rearranged by dragging the cell with the name of the factor attribute heading
a top or a left head. Drag it to the near of another factor attribute to move the head with its values. An
optimal arrangement that fits the width of the window can be get by choosing Sort from the menu
Edit.
2.4.1.4 Consistency
Attributes, which make sense of something goodness allows to compare two rules based on the
goodness of their domain and output. Assumed that a rule with better domain can not have a worse
output, consistency can be defined. A rule is inconsistent with a rule set, if it has a rule with better
domain and worse output, or a rule with worse domain and better output. A rule may be too good or
too bad for this set respectively and if it is too good and too bad at the same time it is called rule in a
crossfire.
On this base
newly added rules can be checked (Flag Inconsistent Rules (R-2.4.1.5)),
R-53
too bad, too good rules or rules in crossfire can be searched for (Examine (R-2.4.3)),
consistent advice can be shown in the status bar or in the empty 2D cells (Show Plausible
Advice in Customize (R-2.4.1.1)),
or an output for uncovered domains can be estimated (Cases Accept Advice (R-2.2.4.2)).
2.4.1.5 Flag Inconsistent Rules
Switch Flag Inconsistent Rules on by Customize (R-2.4.1.1). Change the output (R-2.4.2.5) of a
rule. If small faces appear on the right of some cells, it means that you have entered an inconsistency
into the rule set.
In such cases the rule last entered has a yellow face. The rules that are in contradiction to it have
white ones. If more than one rule are changed, only one of them is yellow, the other ones are grey, but
the corresponding rules are not signed until they are selected by the cursor, by the time they will be
yellow.
Too bad rules (R-2.4.3.7) are signed by a crying, too good (R-2.4.3.8) ones by a smiling, and rules in
crossfire (R-2.4.3.9) are signed by a face with waving mouth.
In order to hide inconsistency flags you can

alter the output of the rule last entered or those being in contradiction to it,
set value ordering (R-2.1.2.8.1) of the obscured factor attribute to Nominal,
set value ordering (R-2.1.2.8.1) of the dependent attribute selves to Nominal to switch off
consistency check for this attribute,
or switch Flag Inconsistent Rules off by Customize (R-2.4.1.1).
Advanced: This function of Doctus checks the inconsistency of complex rules, which often
can have extensive active domains. This is why sometimes additional white faces appear on the
domain, which is not in contradiction to the domain of the checked rule. On contrary, Search
Contradiction (R-2.4.3.6) works on domain base.
2.4.2 Edit
2.4.2.1 Insert Rule
Choose Insert Rule from the menu Edit. This dialog is opened:
R-54
Figure R-19: Selecting the domain of the purple rule on Figure R-17.
Select a domain, multiple ranges can be selected by holding down the Ctrl. For full range simply
leave empty the row of the factor. Press OK.
In the Rule list view a new rule is inserted before the current one, with Unknown output. Use the
same command in 2D view or simply change the cell content from the context menu. The new rule will
be inserted on the top of the rule list.
Another way to create a rule is to set a value in the output cell below the last rule of the list. A new
rule, covering the whole domain, is inserted and its output is set.
2.4.2.2 Delete Rule
Switch to the rule list view. Select the rows of rules to delete and choose Delete Rule from the
menu Edit. The selected rules are deleted. The same action in the 2D view will result in setting the
output of the selected rule(s) to Unknown.
2.4.2.3 Move Rules
Select rows of rules to move, grip the selection at its border, drag, and drop it. The rules are
reordered. If the output of the rule set is altered in consequence of overlapping domains, a warning is
displayed.
2.4.2.4 Move Factor
Select column of factor to move, grip the selection at its border, drag, and drop it. The factors are
reordered. Edges of the Rule-Based Graph are updated too.
For other selections the normal text mode move (R-3.6) is active.
2.4.2.5 Set Rule Output
Move cursor to or select multiple cells of rule outputs and choose a value from the context menu.
Rule output can be set to Unknown, Dont care or distributed values too by the same way as case
features (R-2.2.2.2).
In rule list view the output can also be altered if a whole row is selected.
R-55
In 2D view the selection is transformed into rule domains. If it is needed, new rules are inserted
before the first one with the selected output, otherwise only the output is changed for the appropriate
rules. This action may also result in multiple rules in the rule list.
2.4.2.6 Set Range
Switch to the rule list view. Move cursor to or select cells of a factor. Choose Set Range to from
the context menu. Select a value from the submenu. The selected ranges are set to a range with that
value only.
2.4.2.7 Split Range
Switch to the rule list view. Move cursor to or select cells of a factor. Choose Split at from the
context menu. Select a value from the submenu. The selected ranges are split into two ranges. The
second range starts with the selected value.
2.4.2.8 Divide Range
Switch to the rule list view. Move cursor to or select cells of one or more factors and/or one or
more rules. Choose Divide Range from the menu Edit.
Figure R-20: The effect of Divide Range on all factors and rules of Figure R-21.
The selected ranges of selected rules are divided into single values.
2.4.2.9 Assemble Rules
Choose Assemble Rules from the menu Edit.
Figure R-21: The effect of Assemble Rules on Figure R-20.
R-56
The number of rules is reduced by linking them into complex rules with larger domain.
2.4.2.10 Extend Value Ranges
Choose Extend Value Ranges from the menu Edit.
Figure R-22: The effect of Extend Value Ranges on Figure R-21.
The domains of some rules are extended, so they are more general without altering the rule output.
2.4.2.11 Purify Rule Set
Choose Purify Rule Set from the menu Edit. The domains of the rules are divided (R-2.4.2.8)
into elementary rules, hidden rules are deleted (R-2.4.2.2), the rest is assembled (R-2.4.2.9) and
extended (R-2.4.2.10). Finally the resulted complex rules are sorted (R-2.4.2.12).
Figure R-23: The rule set of Figure R-16 after that its rules are divided. The result is seen on Figure R-22.
Tip: The rule set can be enlarged and confused for a human reader by intensive output
alteration in 2D. Use this function to resolve this situation.
2.4.2.12 Sort
Switch to the rule list view. Select columns to sort by in precedence order, and choose Sort from
the menu Edit. Rules are sorted by values or range starts of the selected output or factors
respectively. If the output of the rule set is altered in consequence of overlapping domains, a warning is
displayed.
R-57
2.4.3 Examine
2.4.3.1 Search the Same Domain on the Other View
Move cursor to or select more rules on the list view. Change to 2D, and the active domains of the
selected rules are selected.
Move cursor to or select multiple cells of rule outputs on 2D. Change to list view, and the rules with
the selected active domain are selected.
Choose Value Combination from the menu Search. This dialog is opened:
Figure R-24: Selecting rules on the domain of mean Finance and with the output suggested Tender. Three cells will be
selected on Figure R-17.
Select the required domain and output. Select multiple ranges with holding down Ctrl. If a factors
range or the output does not matter, leave its row clear. Press OK.
The matched rules are selected either on the rule list view or on the 2D view.
2.4.3.3 Search Hidden Rules
Switch to the rule list view. Choose Hidden Rules from the menu Search. Rules without active
domain, namely totally hidden rules and rules with an empty range are selected. They can be deleted
without altering the output of the rule set.
2.4.3.4 Search Multiple Covered Domains
Switch to the 2D view. Choose Range Intersection from the menu Search. Multiple covered
domain is selected.
R-58
Tip: Use this action to check whether the rule set is single layered.
2.4.3.5 Search Active Domain
Switch to the 2D view. Move cursor to the domain of a complex rule. Choose Active Domain
from the menu Search. The active domain of the rule is selected.
2.4.3.6 Search Contradiction
Switch to the 2D view. Move cursor to a cell of the domain. Choose Contradiction from the menu
Search. The domain is selected that is in contradiction to it in the sense of Consistency (R-2.4.1.4).
Advanced: This action differs from Flag Inconsistent Rules (R-2.4.1.5) in that it works
on pure domain base, so the extensive active domains of complex rules do not affect the result.
2.4.3.7 Search Too Bad Rules
Choose Too Bad Rules from the menu Search. The rules are selected that have another rule with
worse (R-2.4.1.4) domain and better output.
2.4.3.8 Search Too Good Rules
Choose Too Good Rules from the menu Search. The rules are selected that have another rule
with better (R-2.4.1.4) domain and worse output.
2.4.3.9 Search Rules in Crossfire
Choose Rules in Crossfire from the menu Search. It will select the rules that have other rules
with both worse (R-2.4.1.4) domain and better output and better domain and worse output.
2.4.4 Manage
The Rule-Based Reasoning can be started by

Cases Deductive Reasoning (R-2.2.4.1) or
Rule-Based Graph Deductive Reasoning (R-2.3.4.1).
2.5 Case-Based Graph
The Case-Based Graph depicts the result of the Case-Based Reasoning. The Case Based Reasoning
produces new rules for the benchmark attribute based on case features of other attributes.
R-59
Figure R-25: Attributes of a knowledge base about the ability of employees. It has no rules on the Rules pane, but it has
well described cases.
The produced rules can be read by walking through the graph from the top to a leaf. Check the
attribute seen on the node, and choose the way with the matching label.
Figure R-26: The Case-Based Graph that is generated from the cases of Figure R-25.
Example: How can be Ability evaluated? If Dominant knowledge is intuition and

Motivation is biological then Ability the benchmark is expert.
2.5.1 View
2.5.1.1 Customize
R-60
Figure R-27: Customize Case-Based Graph pane
This dialog controls the appearance of the Case-Based Graph pane. Following switches are
supported:
Horizontal Graph is mirrored that the root node is on the left of the window, and the
branches are on the right of it.
Condensed Nodes with more than two leafs can take a wide area of the screen. Check this flag
to arrange such leafs into two rows to reduce total width of the graph (in horizontal view it
will arrange the leafs into two columns to reduce the hight).
2.5.1.2 Information
Each node of the graph represents a set of cases. The root node (on the top) represents all cases.
An attribute is chosen for each internal node, which is showed in the rounded rectangle. The values
of it are distributed into more branches, which are indicated in rectangular labels. Cases of the node are
distributed too into the leaf nodes based on its features of the chosen attribute.
Leaf nodes has no branches based on an attribute, so instead of it the case features for the
benchmark attribute are shown.
To view the attributes that can be chosen (R-2.5.1.2.1), the case set of the node (R-2.5.1.2.2), or
cases features for the benchmark attribute (R-2.5.1.2.3), select a node and choose Information from
the menu View. A sheet is opened with the following pages:
2.5.1.2.1 Informativity
R-61
Figure R-28: Attributes that can be chosen for branching.
Cases of the node are distributed into branches by its case features for the chosen attribute. This
distribution can be successful only if case features have some dependence on the benchmark. The
measurement of this dependence is called informativity. This table lists the informativities and densities
of the most informative value distribution for each informative attributes.
Maths: How informativity and density are determined, see Appendix Determining
Informativity (R-7.4).
2.5.1.2.2 Cases
Figure R-29: Cases of the node.
This page lists following properties of all cases in the current node:
Case The name of the case.
Weight It can be less than one for distributed case features.
Attribute name The case feature for a chosen attribute. The branching attribute is shown by
default, but the context menu can be used on this column to choose another informative
attribute to show.
Benchmark attribute The case feature for the benchmark attribute.
2.5.1.2.3 Statistics
R-62
Figure R-30: The distribution of the benchmark case features.
The distribution of the case features by the benchmark attribute is shown on this page.
2.5.2 Edit
2.5.2.1 Prune
Select a internal node and choose Prune from the Edit or context menu. All branches are cut
off. All values are shown that occur as a case feature for the benchmark attribute in this node.
2.5.2.2 Branch
Select a leaf node with an informative attribute (R-2.5.1.2.1) and choose Branch from the context
menu. The most informative attribute is chosen as branching attribute and the value distribution with
the highest informativity is used to create branches and leafs. This action can be also performed on the
leafs containing multiple case features for the benchmark attribute.
2.5.2.3 Select Attribute
Select a node with an informative attribute (R-2.5.1.2.1) and choose Select Attribute from the
context menu to change the informative attribute for that node. Select an informative attribute from
the submenu. Branches are grown in the same manner as by Branch (R-2.5.2.2).
2.5.3 Examine
2.5.3.1 Classify
Choose Export from the menu File. On the Advanced tab select the template
CBRclassify.txt.dkt from ExportTemplates\report folder of the Doctus installation. Press OK and
Next. Answer the questions about your case to walk through the nodes of Case-Based Graph. The
last page of the Export Wizard shows the classification of your case. Your answers and the result are
stored into the exported output file.
2.5.4 Manage
2.5.4.1 Generate a Case-Based Graph
Choose Inductive Reasoning from the menu Knowledge Management. The Case-Based Graph is
generated recursively.
R-63
The generation is controlled by following pages of the opened sheet:
2.5.4.1.1 Branching Method
Figure R-31: Generate Case-Based Graph Branching Method
Following options are allowed:

Polar Values are always distributed into two subsets.
Heuristic Values are distributed into two subsets at first, then one of the subsets is divided
on, if the informativity grows by it. And so on.
Efficient All value combinations with arbitrary number of branches are searched through
exhaustively.
Ordered Switch this checkbox on to assure that all branches have adjoining values only. If it is
off, leaky branches can be created too. It can be configured for each attribute (R-2.1.2.8.2)
separately.
Exclude dependent Attributes For mixed-mode knowledge bases, when both Rule- and Case-
Based Reasoning are used, the dependent attributes can be disabled to be chosen as
informative attribute.
2.5.4.1.2 Benchmark
R-64
Figure R-32: Generate Case-Based Graph Benchmark
The benchmark attribute can be selected on this page.
2.5.4.1.3 Thresholds
Figure R-33: Generate Case-Based Graph Thresholds
Three conditions are defined to limit the extension of the automatic generated Case-Based Graph.
Following values are the parameters of this conditions:
Minimal Informativity Density The branching is stopped for a node, if density of the most
informative attribute does not reach this limit.
Maximal Majority of a Benchmark Value The branching is stopped for a node, if the rate of the
most frequently occurred benchmark case feature reaches this limit.
Minimal Weight of Cases in Node The branching is stopped for a node, if the relative weight of
cases falls below this limit.
2.5.4.1.4 Learning Cases
R-65
Figure R-34: Generate Case-Based Graph Learning Cases
The case set of the root node can be reduced on this page. The selection button limits the cases to
the ones selected on the Cases pane, the other buttons are self-explanatory.
2.5.4.2 Automatic Regenerate
Switch this flag on in the Knowledge Management menu to make Doctus regenerate the Case-
Based Graph after each action that possibly alters it.
2.5.4.3 Extract Case-Based Rule Graph
Create a Case-Based Graph and choose Extract Rules from the menu Knowledge Management.
Figure R-35: The Case-Based Rule Graph created from Figure R-26.
The rules of the Case-Based Graph are converted into a single level Rule-Based Graph. Only the
relevant attributes (which appear in the Case-Based Graph) are transmitted. The resulted new
knowledge base can be used for Rule-Based Reasoning and will give the same results as its parent case-
based knowledge base.
R-66
3 Navigation in Tables
3.1 Moving Cursor
The usual keys can be used: Home and End horizontally, Ctrl-Home and Ctrl-End vertically
jump to the appropriate edge of table. Ctrl-Left and Ctrl-Right are for horizontal paging. With the
mouse, use scrollbars and left-click to move cursor.
3.2 Context Menu
Frequently used functions can be called from context (right-click) menu, the content of which also
depends on the selected item.
3.3 Entering Text
Text cells can be overwritten by typing, edited by doubleclicking it or pressing F2. The text
entered is validated only by leaving cell or by pressing Enter. Esc restores the cell to text before
editing. There is also an undo (R-5.1.2.1) function.
3.4 Selection
To select a rectangular field of cells, Shift can be used. The mouse is used as usual. Multiple cells
can be selected by holding down left button. A click on the head selects the entire column, row or
table. Ctrl serves to make multiple selections.
3.5 Inserting and Deleting
Edit.Insert (R-5.1.2.7) inserts new attribute, value, case or rule. The text of the menu shows which
of them is currently available. The position of the cursor determines the position of the new item.
Deletion is possible with Edit.Delete (R-5.1.2.8) or pressing Del, following the selection of the
appropriate cell(s), row(s) or column(s) as demanded.
3.6 Moving Cells
If a single area is selected it can be dragged and dropped using the mouse. Cells can be copied by
holding Ctrl down. The usual cut, copy and paste functions are also available.
If the selected area is a row or a column, moving brings about their reordering:
Attributes Move Attribute (R-2.1.2.7)
Cases Move Attribute (R-2.2.2.8)
R-67
Cases Move Case (R-2.2.2.7)

Rules Move Rules (R-2.4.2.3)
Rules Move Factor (R-2.4.2.4)
3.7 Sizing Cells
By clicking the head on the edge of the cell, it is possible to alter the width of a column, or the
height of a row. If there are selected cells, their size will be changed. Double clicking resizes the cell to
fit the text.
4 Knowledge Export & Import
4.1 Knowledge Export
A knowledge base can be exported into various forms of computer programs. Some of these
perform reasoning, though the Dont care and the distributed case features and rule outputs are not
supported. Some of the exported programs may include a feature of submitting new cases; these are
then called Knowledge Acquisitors, as the acquired cases can be further processed in the knowledge
base. There are two types of Knowledge Acquisitor one sends the new cases in form of encoded e-
mails (R-4.2.4) to a given e-mail address, the other saves them at a given URL as URL encoded cases
(R-4.2.5), i.e. a log file with *.dku extension (R-6.3.5) downloaded from the URL. The Knowledge
Import modul can read these cases into the knowledge base from which the Knowledge Acquisitor was
exported, other knowledge bases cannot interpret the submitted cases.
To start exporting choose Export from the menu File. A dialog is opened where we can choose
from the following forms of export:
4.1.1 HTML
Figure R-36: Exporting into HTML pages
R-68
A HTML page is generated which is able to acquire, reason and/or send features of a case.
Following types are implemented:
Single Deductive The case features can be selected (R-2.2.4.1) on the same page for all input
attributes.
Multiple Deductive The case features can be selected (R-2.2.4.1) page by page for each input
attribute.
Multiple Inductive A value of each relevant informative attribute (R-2.5) can be selected page
by page.
Enter title of page into Title and select Language. Choose Color scheme. Select knowledge
processing mode at Finish buttons as follows:
Reason Set this checkbox to put Conclusion attribute only or All dependent attributes on
the page with a Reason button. Pressing this button activates the rules that are exported in
javascript into the page and shows the outputs of the reasoning on the page.
Send Set this checkbox to generate a Send button, which sends the acquisited case features
to the given e-mail address. These e-mailed cases can be then imported (R-4.2.4) into the
knowledge base.
Press OK and select file to store generated page to.
4.1.2 Graphs
Figure R-37: Exporting pictures of graphs
The Rule-Based Graph and the Case-Based Graph can be exported into different graphical formats:
Office Drawing Shapes Graph is stored on the clipboard in Microsoft Office Drawing Shapes
format. It can be inserted into various Microsoft Office documents or into documents of
other applications supporting this format.
GIF Graph is stored in a GIF file.
DIB Graph is stored in a Device Independent Bitmap file.
R-69
4.1.3 Tables
Figure R-38: Exporting tables of Doctus
The content of different tables of Doctus can be exported in the following formats:
Excel Workbook Tables are stored in different sheets of a Microsoft Excel Workbook.
GIF The screenshot of the whole tables are stored in postfixed GIF files.
DIB The screenshot of the whole tables are stored in postfixed Device Independent
Bitmap files.
Following tables can be selected to export:

Attributes The Attributes pane.
Cases The Cases pane.
Rules The Rules pane in rule List, 2D, or in its original Default view mode for the
Current attribute only or for All dependent attributes.
Case-Based Node The information window of the Case-Based Graph, independently selectable
the Informativity, the Cases or the Statistics page, for the Current only or for All
nodes.
Cluster configuration The cluster configuration table for the Current attribute only or for All
numeric attributes.
Press All to select all tables or press None to deselect all tables.
R-70
4.1.4 Advanced
Figure R-39: Exporting by an export template
The current knowledge base is exported by a given export template. Select template (R-6.3.4) and
output files, and press OK. The Export Wizard queries all relevant template parameters, and
generates the exported knowledge. The export templates delivered by Doctus are detailed in Appendix
Export Templates (R-7.2).
Advanced: Build your own export template based on the delivered ones. A syntax sheet can
be found in the installation folder of Doctus at ExportTemplates/ReadMe.txt.
4.2 Knowledge Import
Case features can be acquired directly from an external data source. By doing so, the values of the
attributes can be automatically acquired and the numerical data is clustered. Attributes can be linked to
database or table columns using the Link Wizard (R-4.2.6). This set-up can later be modified here or at
Attributes Attribute Properties (R-2.1.2.8). For importing choose Import from the menu File. A
dialog is opened where we can choose from the following import sources (some of them are displayed
in advanced mode only):
R-71
4.2.1 Excel Workbook
Figure R-40: Importing from an Excel Workbook
Select this page if case features are stored in an existing Microsoft Excel Workbook (*.xls). The
cases have to be arranged into rows, where the first row contains the column names, to which
attributes can be linked. Excel Workbooks can be selected from the list, entered manually or searched
after pressing Browse.
4.2.2 Microsoft Query (in advanced mode only)
Figure R-41: Importing from a database query by Microsoft Query
Select this page to query case features from more database tables with logical dependencies.
Microsoft Query allows the graphical creation of complex database queries. An existing query file can
be selected from the list, entered manually or searched after pressing Browse. Pressing New a new
query can be created, which can be later saved into a file.
R-72
4.2.3 ODBC (in advanced mode only)
Figure R-42: Importing from an ODBC database query
Select this page to import case features directly from independent database tables or data source for
which direct Doctus driver does not exist. The knowledge base can be connected to the external data
source using built-in ODBC (Open Database Connectivity) driver of Windows. A driver (*.dsn) can be
selected from the list, entered manually or searched after pressing Browse. Pressing New a new
driver can be created, which can be later saved into a file.
4.2.4 Mailbox
Figure R-43: Importing from a mailbox
Select this pane to read case features from encoded e-mails sent by a Knowledge Acquisitor (R-4)
page generated from the current knowledge base. One e-mail message contains features of one case.
Since the page has been generated from the same knowledge base, no further attribute linking is needed
and the cases cannot be interpreted by another knowledge base. This import feature starts the default e-
mailing software, which may open an additional dialog box to select the e-mail account and/or confirm
R-73
its opening. Doctus searches the Inbox of the selected account for encoded e-mail messages containing
the case features. As there is no need for linking the attributes the Link Wizard is disabled; so once the
source of import is selected only choose the clustering options (R-4.2.7) and start the import (R-4.2.8).
4.2.5 URL Encoded Cases (in advanced mode only)
Figure R-44: Importing from a file with URL Encoded Cases
Select this pane to read features of URL encoded cases created by Knowledge Acquisitor (R-4)
exported from the same knowledge base. This Knowledge Acquisitor is a webserver application (R-
4.1.4), which writes all the acquired cases into a single *.dku file (R-6.3.5) at a given URL. To import
these cases, we should download the file from the URL and select it using the Browse button. Since
the webserver application has been generated from the same knowledge base, no further attribute
linking is needed and the cases cannot be interpreted by another knowledge base. As there is no need
for linking the attributes the Link Wizard is disabled; so once the source of import is selected only
choose the clustering options (R-4.2.7) and start the import (R-4.2.8).
4.2.6 Link Wizard
When importing case features from a third-party external database (i.e. Excel Workbook (R-4.2.1),
Microsoft Query (R-4.2.2) and ODBC source (R-4.2.3)) we have to relate the attributes to link to the
corresponding table columns (R-2.1.2.8.5) and to configure the auto values (R-2.1.2.8.3). This is
supported by the Link Wizard.
Select Attributes to link in the list box, Link Wizard will prompt each of them to be set up. You
can select attributes Without values, the Already linked ones, or All by pressing the appropriate
button.
The Link Wizard is opened for the appropriate data sources, if at least one attribute is selected to
link:
R-74
Figure R-45: Linking the table column done_works to the attribute Done works.
Select the required number of clusters in the list box Cluster, or select Text for flexible values
(R-2.1.2.8.3) or None for constant ones (R-2.1.2.8.3). For this last option the undefined values are
replaced with Unknown. Then select Table and Column of the data source to link attribute to.
Case name combo box is for selecting the column that identifies the table rows, i.e. the cases. Press
Next or double click Cluster or Column to establish this link and get next attribute, which is
selected, automatically in the combo box Attribute to link. For the last attribute press Finish to
finalize the links.
4.2.7 Clustering options
The clustering options are available for all import types, these can be set using the Clustering radio
buttons. Choose:
Restart to restart the cluster analysis. Non-renamed clusters (i.e. the values beginning with
/) will be deleted, and the requested number of clusters will be built, While the renamed
clusters will keep their new names but the clusters indicated by these names will be newly
built.
Continue to continue cluster analysis with additional numeric data.
Do not alter to disable cluster analysis. Data will be classified into the existing clusters.
4.2.8 Starting Import
The following finishing options are available in the import dialog:

Cancel Closes the dialog discarding the settings.
Link only Starts the Link Wizard not followed by importing case features. This option is
disabled for import from mailbox and URL encoded cases.
Select cases to import Opens a new dialog with a list to select a subset of cases from the data
source to import. If there is no link established but attributes are selected for linking, this
button starts the Link Wizard, which is followed by starting the import.
Import all Imports all cases from the data source. If there is no link established but attributes
are selected for linking, this button starts the Link Wizard, which is followed by starting the
import.
R-75
5 User Controls
5.1 Main Menu
5.1.1 File
5.1.1.1 File New
Opens an empty knowledge base.
5.1.1.2 File Open...
Opens a saved knowledge base (R-6.3.1).
5.1.1.3 File Close
Closes the current knowledge base.
5.1.1.4 File Save
Saves the current knowledge base and the display styles of the current window into a file (R-6.3.1).
5.1.1.5 File Save As...
Saves the current knowledge base to a selected file (R-6.3.1).
5.1.1.6 File Run Batch
Executes a Doctus batch file (R-6.3.2) on the current knowledge base(s).
5.1.1.7 File Properties
The following properties are shown:

Version number increases every time by saving the file
Knowledge base description can be edited by this dialog
Statistics various numbers describing the knowledge base
This window also opens in Windows Explorer on Details pane of Properties for dkb files (R-
6.3.1).
5.1.1.8 File Export
See Knowledge Export (R-4.1).
5.1.1.9 File Import
See Knowledge Import (R-4.2).
R-76
5.1.1.10 File Exit
Exits Doctus.
5.1.2 Edit
5.1.2.1 Edit Undo
Undoes last action.
5.1.2.2 Edit Redo
Redoes the action last undone.
5.1.2.3 Edit Cut
Copies cell selection (R-5.1.2.4) into the clipboard and deletes (R-5.1.2.8) it.
5.1.2.4 Edit Copy
Copies the content of selected cells into the clipboard.
Technical: The clipboard format that is internally used to copy cells is a Tab delimited
textual form. It is supported by Microsoft Excel and many other spreadheet applications and
text editors, so cells can be copied both ways.
5.1.2.5 Edit Paste
Writes over cells with the content of the clipboard. No insert is performed.
5.1.2.6 Edit Paste External Data
See Cases Paste External Data (R-2.2.2.9).
5.1.2.7 Edit Insert
Context-dependent command. Related actions are:

Attributes Insert Attribute (R-2.1.2.3)
Attributes Insert Value (R-2.1.2.4)
Cases Insert Case (R-2.2.2.5)
Rules Insert Rule (R-2.4.2.1)
5.1.2.8 Edit Delete

Attributes Delete Attribute (R-2.1.2.5)
Attributes Delete Value (R-2.1.2.6)
Cases Delete Case (R-2.2.2.6)
Rules Delete Rule (R-2.4.2.2)
R-77
5.1.2.9 Edit Set to Dont Care
Sets case feature or rule output to Dont care. See also Cases Enter case features (R-2.2.2.2) and
Rules Set Rule Output (R-2.4.2.5).
5.1.2.10 Edit Distribute
Distributes case feature or rule output. See Cases Enter distributed case feature (R-2.2.2.3).
5.1.2.11 Edit Divide Range
See Rules Divide Range (R-2.4.2.8).
5.1.2.12 Edit Assemble Rules
See Rules Assemble Rules (R-2.4.2.9).
5.1.2.13 Extend Value Ranges
See Rules Extend Value Ranges (R-2.4.2.10).
5.1.2.14 Edit Purify Rule Set
See Rules Purify Rule Set (R-2.4.2.11).
5.1.2.15 Edit Sort
See Cases Sort (R-2.2.2.10) and Rules Sort (R-2.4.2.12).
5.1.2.16 Edit Attribute Properties
See Attributes Attribute Properties (R-2.1.2.8).
5.1.2.17 Edit Set to Decision Attribute
See Attributes Set to Decision Attribute (R-2.1.2.9).
5.1.2.18 Edit Cluster Configuration
See Attributes Cluster Configuration (R-2.1.2.8.4).
5.1.2.19 Edit Prune
See Rule-Based Graph Prune (R-2.3.2.3) and Case-Based Graph Prune (R-2.5.2.1).
5.1.3 View
5.1.3.1 View Toolbar
The toolbar is showed or hidden. See Toolbar (R-5.3).
R-78
5.1.3.2 View Status Bar
The status bar is showed or hidden.
5.1.3.3 View Description & Information

Attributes Description (R-2.1.1.1)
Cases Description (R-2.2.1.1)
Case-Based Graph Information (R-2.5.1.2)
5.1.3.4 View Fonts
The display fonts can be set up by this dialog:
Figure R-46: Font Settings
Choose Filter to reduce the size of Display item list. Select Display item(s) to set up, then
choose Character set, Typeface, Size, and optionally Bold or Italic. Choose another Display
item to set up. Press OK to accept changes.
5.1.3.5 View Customize

Attributes Customize (R-2.1.1.2)
Cases Customize (R-2.2.1.2)
Rule-Based Graph Customize (R-2.3.1.1)
Rules Customize (R-2.4.1.1)
Case-Based Graph Customize (R-2.5.1.1)
R-79
5.1.3.6 View Style Organizer
This option is available in advanced mode only. The display fonts and cell sizes are organized by this
dialog:
Figure R-47: Organize Display Styles
Choose Filter to reduce the size of Display items to change list. Select Display items to change
and Properties to change, then choose one of the following actions:
Reset Resets selected properties of selected display items to the hardcoded default.
Load from File Loads selected properties of selected display items from a file (R-6.3.3).
Load from a window Loads selected properties of selected display items from another child
window.
Save to Saves selected properties of selected display items from the current window, and the
other ones as default into a file (R-6.3.3).
Press OK to accept changes that affect the current window.
5.1.4 Search
See Cases Search Value Combination (R-2.2.3.1) and Rules Search Value Combination (R-
2.4.3.2).
5.1.4.2 Search Unused Value
See Attributes Search Unused Value (R-2.1.3.1).
See Attributes Search Factors (R-2.1.3.2) and Cases Search Factors (R-2.2.3.2).
5.1.4.4 Search Explanation
See Cases Search Explanation (R-2.2.3.3) and Rule-Based Graph Explain (R-2.3.3.1).
R-80
5.1.4.5 Search Hidden Rules
See Rules Search Hidden Rules (R-2.4.3.3).
5.1.4.6 Search Multiple Covered Domain
See Rules Search Multiple Covered Domains (R-2.4.3.4).
5.1.4.7 Search Active Domain
See Rules Search Active Domain (R-2.4.3.5).
5.1.4.8 Search Contradiction
See Rules Search Contradiction (R-2.4.3.6).
5.1.4.9 Search Too Bad Rules
See Rules Search Too Bad Rules (R-2.4.3.7).
5.1.4.10 Search Too Good Rules
See Rules Search Too Good Rules (R-2.4.3.8).
5.1.4.11 Search Rules in Crossfire
See Rules Search Rules in Crossfire (R-2.4.3.9).
5.1.5 Knowledge Management
5.1.5.1 Knowledge Management Deductive Reasoning
See Cases Deductive Reasoning (R-2.2.4.1) and Rule-Based Graph Deductive Reasoning (R-
2.3.4.1).
5.1.5.2 Knowledge Management Accept Advice
See Cases Accept Advice (R-2.2.4.2) and Rule-Based Graph Accept Advice (R-2.3.4.2).
5.1.5.3 Knowledge Management Inductive Reasoning
See Case-Based Graph Generate a Case-Based Graph (R-2.5.4.1).
5.1.5.4 Knowledge Management Automatic Regenerate
See Case-Based Graph Automatic Regenerate (R-2.5.4.2).
5.1.5.5 Knowledge Management Branch
See Case-Based Graph Branch (R-2.5.2.2).
R-81
5.1.5.6 Knowledge Management Extract Rules
See Case-Based Graph Extract Case-Based Rule Graph (R-2.5.4.3).
5.1.5.7 Knowledge Management Terminate Process
Terminates Knowledge Management Inductive Reasoning (R-5.1.5.3) manually. Use it if the

reasoning seems to get into infinite cycle or if the estimated timing is to long.
5.1.6 Window
5.1.6.1 Window New Window
Opens a new window for the current knowledge base. It has an own display style (R-5.1.3.6).
5.1.6.2 Window Split
Opens a new window (R-5.1.6.1) for the current knowledge base and arranges it with the current
one tiled (R-5.1.6.4).
5.1.6.3 Window Cascade
Arranges windows so they overlap.
5.1.6.4 Window Tile
Arranges windows as non-overlapping tiles horizontally.
5.1.6.5 Window Tile Vertical
Arranges windows as non-overlapping tiles vertically.
5.1.6.6 Window Arrange Icons
Arranges icons at the bottom of the window.
5.1.7 Help
5.1.7.1 Help Help
Displays help for current task or command.
5.1.7.2 Help Help Topics
Displays this document as a help.
5.1.7.3 Help Context Help
Switches on context help mode. Select a menu or toolbar item to get help on it.
R-82
5.1.7.4 Help Enter Authorization Key
Figure R-48: Entering authorization key
Opens a dialog to enter authorization key, which enables full operation of Doctus.
5.1.7.5 Help About Doctus...
Opens the about dialog. Running mode and software version is shown.
R-83
5.2 Keyboard Shortcuts
Ctrl+1 View current attribute only on the Cases pane. See Cases
Customize (R-2.2.1.2).
Ctrl+2 Switch between rule list and 2D view on the Rules pane. See
Rules Customize (R-2.4.1.1).
Ctrl+C Main Menu Edit Copy (R-5.1.2.4)
Ctrl+D Main Menu Edit Divide Range (R-5.1.2.11)
Ctrl+E Main Menu Edit Paste External Data (R-5.1.2.6)
Ctrl+F Main Menu View Fonts (R-5.1.3.4)
Ctrl+I Main Menu View Description & Information (R-5.1.3.3)
Ctrl+N Main Menu File New (R-5.1.1.1)
Ctrl+O Main Menu File Open... (R-5.1.1.2)
Ctrl+P Main Menu Edit Attribute Properties (R-5.1.2.16)
Ctrl+S Main Menu File Save (R-5.1.1.4)
Ctrl+T Main Menu View Customize (R-5.1.3.5)
Ctrl+U Main Menu Edit Cluster Configuration (R-5.1.2.18)
Ctrl+V Main Menu Edit Paste (R-5.1.2.5)
Ctrl+X Main Menu Edit Cut (R-5.1.2.3)
Ctrl+Y Main Menu Edit Redo (R-5.1.2.2)
Ctrl+Z Main Menu Edit Undo (R-5.1.2.1)
F1 Main Menu Help Help (R-5.1.7.1)
Shift+F1 Main Menu Help Context Help (R-5.1.7.3)
Shift+F2 Edit cell text.
Ctrl+F6 Switch to the next window.
Ctrl+Shift+F6 Switch to the previous window.
Alt+Backspace Main Menu Edit Undo (R-5.1.2.1)
Shift+Del Main Menu Edit Cut (R-5.1.2.3)
Ins Main Menu Edit Insert (R-5.1.2.7)
Ctrl+Ins Main Menu Edit Copy (R-5.1.2.4)
Shift+Ins Main Menu Edit Paste (R-5.1.2.5)
Ctrl+PgDn Switch to the next pane.
Ctrl+PgUp Switch to the previous pane.
Ctrl+Down Go to the next attribute.
Ctrl+Up Go to the previous attribute.
Ctrl+Enter Go to the start of next line.
Table R-1: Keyboard Shortcuts
R-84
5.3 Toolbar
Main Menu File New (R-5.1.1.1)

Main Menu File Open... (R-5.1.1.2)
Main Menu File Save (R-5.1.1.4)
Main Menu File Import (R-5.1.1.9)
Main Menu File Export (R-5.1.1.8)
Main Menu Edit Delete (R-5.1.2.8)
Main Menu Edit Cut (R-5.1.2.3)
Main Menu Edit Copy (R-5.1.2.4)
Main Menu Edit Paste (R-5.1.2.5)
Main Menu Edit Undo (R-5.1.2.1)
Main Menu Edit Redo (R-5.1.2.2)
Main Menu Edit Divide Range (R-5.1.2.11)
Main Menu Edit Purify Rule Set (R-5.1.2.14)
Main Menu Edit Sort (R-5.1.2.15)
Main Menu Search Too Bad Rules (R-5.1.4.9)
Main Menu Search Too Good Rules (R-5.1.4.10)
Main Menu Search Rules in Crossfire (R-5.1.4.11)
Main Menu Knowledge Management Deductive Reasoning (R-5.1.5.1)
Main Menu Knowledge Management Inductive Reasoning (R-5.1.5.3)
Switch between horizontal and vertical view (R-5.1.3.5)
Switch between the rule list and 2 dimensional view on Rules pane (R-2.4.1.1)
Main Menu View Customize (R-5.1.3.5)
Main Menu View Description & Information (R-5.1.3.3)
Main Menu Help Help (R-5.1.7.1)
Main Menu Help Context Help (R-5.1.7.3)
Zoom graph
Table R-2: The Toolbar of Doctus
6 Operating & Maintenance
This chapter contains some technical information about Doctus.
6.1 Installation
Start the setup using the file received on CD or downloaded from the following location:
http://www.doctus.info/download#exe. Select installation folder in the dialog box. Check to start
Doctus at the end of the installation process. Press Unzip to start installation.
At the first run of Doctus icons can be created in the Start menu, in the Programs menu and/or
on the desktop to start Doctus.
R-85
6.2 Demo and Collector modes
Figure R-49: Choosing running mode
Until Doctus is authorized (R-5.1.7.4) it asks to choose the running mode at startup. Demo mode
can handle only 7 attributes and 14 cases. Collector mode is not capable of knowledge processing.
Authorize opens a dialog to enter authorization key.
6.3 Doctus File Types
Some Windows file extensions are registered for Doctus. The files of these types can be opened by
Doctus by double clicking them in the Explorer, by choosing Open from the context menu, or by
dragging them into Doctus.
6.3.1 Doctus Knowledge Base (*.dkb)
Stores a knowledge base with a display style. If opened, it is loaded. The property sheet for this type
that is opened by the Explorer is enhanced by a new Details (R-5.1.1.7) page. See also Main Menu
File Open... (R-5.1.1.2), Main Menu File Save (R-5.1.1.4) and Main Menu File Save As... (R-
5.1.1.5).
6.3.2 Doctus Batch (*.dka)
R-86
Stores textual Doctus commands (R-7.3). It can be created and edited by a text editor; allowed
commands are detailed in Appendix Batch Commands (R-7.3). If opened, it is executed on the open
knowledge base(s). See also Main Menu File Run Batch (R-5.1.1.6).
6.3.3 Doctus Display Style (*.dks)
Stores cell sizes and font settings. If opened, it is loaded into the current knowledge base. See Main
Menu View Customize (R-5.1.3.5).
6.3.4 Doctus Export Template (*.dkt)
Stores numerous types of programs or documents with special control codes that allows to generate
the knowledge base (or some parts of it) into a resulting program or document file. If opened, the
generation is started from the current knowledge base. Some templates (R-7.2) are delivered with
Doctus. See Knowledge Export Advanced (R-4.1.4).
6.3.5 URL Encoded Cases (*.dku)
It is a log file that stores URL encoded case features written by a webserver application of Doctus. If
downloaded and opened, the case features are imported into the current knowledge base. See
Knowledge Import URL Encoded Cases (R-4.2.5) and Appendix Web Server Applications (R-
7.2.5).
6.4 Advanced Mode
Following advanced features of Doctus are disabled by default:

Run batch (R-5.1.1.6)
Import from Mailbox (R-4.2.4) & URL Encoded Cases (R-4.2.5)
Export with Advanced (R-4.1.4) templates
Main Menu View Style Organizer (R-5.1.3.6)
They can be enabled by merging the file AdvancedMode.reg from the installation folder into the
registry, i.e. by executing it.
6.5 Starting Doctus
Use one of the following ways:

Start doctus.exe from the installation folder.
Use a shortcut created at first run by Doctus or created by you.
R-87
Open a file of a Doctus File Type (R-6.3) from the Explorer.
6.6 Command Line Arguments
Supply a Doctus Knowledge Base (R-6.3.1) to open it or a Doctus Batch (R-6.3.2) to execute it.
6.7 Uninstall
Doctus can be removed using Add/remove programs from the Control Panel. Installed files,
icons and items from the registry will be deleted. The file uninstall.exe and the install folder have to be
removed manually.
7 Appendix
7.1 Samples
7.1.1 Tender (tender.dkb)
This knowledge base evaluates applicants for a tender. Four competitors are assesed, using
25+16+25=66 elementary rules given for 1 decision, two dependent and 4 input attributes.
7.1.2 Tender with Clustering (tender_cluster.dkb)
The same as Tender (R-7.1.1), but Done works, Running works and Price are clustered. Try to
import this (samples/tender_cluster.xls) Excel Workbook.
7.1.3 Ability (ability.dkb)
How can the skills of colleagues be induced from their other attributes? The answer of Doctus is
concluded from the 14 given cases: it is sufficient to examine their Dominant knowledge, Mind,
Motivation and EQ.
7.2 Export Templates
Some export templates (R-6.3.4) are delivered with Doctus to support knowledge export (R-4.1).
R-88
Important note! One do not need the details from this chapter to use the built-in templates of
Doctus. These definitions and descriptions are needed to those developing their own templates based
on the built-in ones, thus even the understanding of this chapter requires significant programming
knowledge.
The export template files are stored in a structured folder system. The basic folder is the
ExportTemplates subfolder of the installation folder of Doctus; within it the subfolders are
hierachically structured according to the export types, containing the template files and the build-in
elements.
7.2.1 Parameter Types
The Export Wizard (R-4.1.4) queries just the parameters that are needed by the generation.
Following types are supported:
Code Type GUI Result

b boolean checkbox 0 or 1
s selection listbox zero based index of selected item
m multiple selection multiple selection list box string of 0s and 1s
r display message static text none
only
t text input edit box input text
f file name selection edit box with Browse full path file name
button
Table R-3: Parameter Types of Export Templates
Advanced: Other syntax elements of the export template language (R-6.3.4) are listed in
ExportTemplates/ReadMe.txt in the installation folder of Doctus.
7.2.2 Desktop Applications
The Rule-Based Reasoning functionality can be exported into standalone computer softwares with
these templates. The starting point can be a Rule-Based Knowledge Base or a Case-Based Graph.
Template Location: ExportTemplates\application
Name Parameters Description

reason.c.dkt s: conclusion only / all The generated c program inputs case features
dependent from stdin and writes conclusion to stdout.
reason.ksh.dkt s: conclusion only / all The generated ksh program inputs case
dependent features from stdin and writes conclusion to
stdout.
Excel.bas.dkt s: horizontal/vertical axis The generated Visual Basic macros create
b: copy cases too sheets as e.g. the Cases sheet with real-time
reasoned dependent attributes.
Table R-4: Export Templates Desktop Applications
7.2.3 Reports
These templates generate no software but data.
Template Location: ExportTemplates\report
R-89

CBRclassify.txt.dkt dynamic Classifies a case during the generation process
based only on its informative features.
Tables.rtf.dkt m: types of tables Writes tables in printable rtf form.
Table R-5: Export Templates Reports
7.2.4 Web Clients
Case features for the input attributes can be given on the web page generated from these templates.
Inputs can be sent and/or used for reasoning by this page or a webserver application (R-7.2.5).
Template Location: ExportTemplates\web\client
Build-in Elements Location: ExportTemplates\web\include

<common> s: language Parameters used by all web clients.
t: title of page
s: style sheet integration
mode
f: style sheet file
<deductive> s: input control (combo box Parameters used by deductive web clients.
/ radio button)
m: on-page reasoning /
webserver / e-mail
s: conclusion only / all
dependent
<deductive on s: output of on-page Parameters used by deductive web clients with
page reasoning> reasoning (combo box / on-page reasoning.
radio button / plain text)
<deductive on t: webserver URL to send to Parameters used by deductive web clients if
server> server is attached.
<deductive send t: e-mail address to send to Parameters used by deductive web clients if
mail> t: mail subject data is sent by mail.
t: acknowledgement page
URL
single_deductive.ht See above. Inputs case features on a single page.
ml.dkt
single_inductive.ht See <common> above. Walks through the nodes of the Case-Based
ml.dkt Graph in hypertext manner.
multi_deductive.ht See above. Inputs case features; on separate page each.
ml.dkt
multi_inductive.ht See <common> above. Walks through the nodes of the Case-Based
ml.dkt Graph page by page; the last page shows the
classification.
Table R-6: Export Templates Web Clients
7.2.5 Web Server Applications
Webserver application that uses the CGI standard to respond http requests of Web Clients (R-7.2.4).
Case features are encoded into http request parameters in the form of
R-90
a<attribute_index>=<case_feature>, Where <attribute_index> is the zero-based index of the

attribute on the Attributes pane, and <case_feature> is 0 for Unknown, else the value index from
1. URL encoded cases (R-6.3.5) are built separated by & from these parameters; each line stores
features of one case.
Template Location: ExportTemplates\web\server
Build-in Elements Location: ExportTemplates\web\include

<common> s: language Parameters used by all webserver applications.
t: title of page
s: style sheet integration
mode
f: style sheet file
t: file name to store URL
encoded cases (R-6.3.5)
save.jsp.dkt See <common>. Java Server Page to store http requests.
save.php3.dkt See <common>. PHP3 script to store http requests.
save.pl.dkt See <common>. Perl script to store http requests.
<reason> <common> Parameters used by reasoner applications.
s: conclusion only / all
dependent
reason.jsp.dkt See <reason>. Java Server Page to store and respond http
requests by a page with reasoning results.
reason.php3.dkt See <reason>. PHP3 script to store and respond http requests
by a page with reasoning results.
reason.pl.dkt See <reason>. Perl script to store and respond http requests
by a page with reasoning results.
Table R-7: Export Templates Web Servers
R-91
7.3 Batch Commands
Cmd Arguments Legal values Action

fo file name \path\*.dkb Main Menu File Open... (R-5.1.1.2)
fs - - Main Menu File Save (R-5.1.1.4)
fa file name \path\*.dkb Main Menu File Save As... (R-5.1.1.5)
fe template name \path\*.dkt Main Menu File Export (R-5.1.1.8)
output file \path\*.*
[parameters...] See Result column.
fi source e: Excel workbook Main Menu File Import (R-5.1.1.9)
q: Microsoft query
o: ODBC
m: mailbox
u: URL encoded
cases
cluster r: restart
c: continue
n: no change
file name *.xls for e
*.qry for q
*.dns for o
*.dku for u
fx - - Main Menu File Exit (R-5.1.1.10)
sa attribute name - Select the attribute.
kr - - Main Menu Knowledge Management
Deductive Reasoning (R-5.1.5.1)
km branching method p: polar Main Menu Knowledge Management
h: heuristic Inductive Reasoning (R-5.1.5.3)
e: efficient
ordered -: not ordered
o: ordered
exclude dependent -: include
x: exclude
benchmark attribute name or *
attribute
ke - - Main Menu Knowledge Management
Extract Rules (R-5.1.5.6)
Table R-8: Batch Commands
7.4 Determining Informativity
Determining informativity (Ib) of attribute b:

1. Let C be the set of cases in a node, a the benchmark, a1an its values, and wa1wan ( )
their rates in set C. The entropy of benchmark in set C is: .
2. Let b1bm be the values of attribute b, is a set of them. Disjoint into not empty subsets
1p, where .
R-92
3. Disjoint C into subsets C1Cp being attribute b of all elements of Cj in j for each j. Let wj be the
weight of Cj in C, where .
4. Then informativity is the entropy gain of disjoining into 1p: .
5. The output of the computing is informativity of the optimal selection: , where k
indicates the possible disjoints of .
Density of it is: , where wC is the number of elements in set C.
7.5 Glossary
2D (2 dimensional) view The rule set is projected into 2 dimensions, where each cell of the table
contains a rule output for the domain defined by the position of the cell. See Rules 2 Dimensional (R-
2.4.1.3).
2 dimensional
Active domain A rule domain in a node of the Rule-Based Graph, which is covered by the rule set
and is not hidden.
Attribute An aspect of evaluation of the cases. See Attributes (R-2.1).
Attribute type Indicates the dependency relations. See Rule-Based Graph (R-2.3). The following
types are distinguished:
1. Single is not a factor and it has no factors.
2. Input is a factor, but it has no further factors.
3. Decision has factors, but it is not a factor.
4. Intermediate or dependent has factors and it is a factor too.
Benchmark attribute The criterion of case classification in the Case-Based Reasoning. See Case-Based
Graph Benchmark (R-2.5.4.1.2).
Case Item evaluated in the knowledge base; e.g. if the knowledge base is used for decision support,
the cases are decision alternatives. See Cases (R-2.2).
Case-Based Graph Graphically displays the result of Case-Based Reasoning. See Case-Based Graph
(R-2.5).
Case-Based Knowledge Base Knowledge base, which uses Case-Based Reasoning; therefore it contains
a Case-Based Graph. See Ability (R-7.1.3).
R-93
Case-Based Reasoning (CBR) The process of inducing rules that classify the cases according to the
values of the benchmark attribute; using informative attributes. See Case-Based Graph Generate a
Case-Based Graph (R-2.5.4.1).
Case-Based Rule Graph When extracting rules form a Case-Based Graph, a new knowledge base is
created with a single-level Rule-Based Graph. For distinction, this graph is called the Case-Based Rule
Graph. See Case-Based Graph Extract Case-Based Rule Graph (R-2.5.4.3).
Case features Values of attributes assigned to a case. See Cases Enter case features (R-2.2.2.2).
Cluster The numeric domain is broken down into parts called clusters. The cases are classified into
these clusters, which will be the case features. See Attributes Cluster Configuration (R-2.1.2.8.4).
Complete rule set A rule set that totally covers the rule domain. See Rules (R-2.4).
Complex rule A rule that covers more than one combination of factor values. See Rules (R-2.4).
Decision attribute IAn attribute that has factors, but it is not a factor of another attribute. Although,
according to this definition, there may be multiple decision attributes only one of them is the logical
conclusion of Rule-Based Reasoning.
Deduction The logic of Rule-Based Reasoning. See Rule-Based Graph Deductive Reasoning (R-
2.3.4.1).
Distributed value A case feature or rule output that is not an exact value of an attribute but can have
different values with different probabilities. See Cases Enter distributed case feature (R-2.2.2.3).
Dont care A case feature or rule output that can have any value of the attribute.
Elementary rule A rule that covers exactly one combination of factor values. See Rules (R-2.4).
Empty range A value range that covers no values. See Rules (R-2.4).
Export Wizard Queries generating parameters required by the chosen export template. See
Knowledge Export Advanced (R-4.1.4).
External source A data source that can be read and interpreted by Doctus. See Knowledge Import
(R-4.2).
Factor An attribute on which another attribute depends. See Rule-Based Graph Connect (R-
2.3.2.1).
Full range A range that covers all values of a factor. Shown as * in the rule list.
Informative attribute Attribute that appears in the Case-Based Graph. See Case-Based Graph
Informativity (R-2.5.1.2.1).
Informativity The measurement of the dependence of an attribute on the benchmark attribute

accoding to the cases in a node. See Determining Informativity (R-7.4).
Knowledge Acquisitor An exported agent capable of acquiring new cases for the knowledge base. See
chapter Knowledge Export Knowledge Export & Import (R-4).
Knowledge base The representation of the knowledge in Doctus. See The Five Panes of Doctus (R-
2).
R-94
Knowledge-Based System (KBS) Computerized system for representing and engineering knowledge.
Consists of a computer program called shell and the knowledge base. See A Knowledge-Based System
(R-1).
Link Wizard A tool for attaching external data sources to attributes and configuring external input.
See Knowledge Import Link Wizard (R-4.2.6).
Reasoning The main knowledge processing function; there are two different types of it: the Rule-
Based Reasoning and the Case-Based Reasoning. See Cases Deductive Reasoning (R-2.2.4.1) and
Case-Based Graph Generate a Case-Based Graph (R-2.5.4.1).
Reduction Extracting informative attributes and rules from the accepted Case-Based Graph. See
Case-Based Graph Extract Case-Based Rule Graph (R-2.5.4.3).
Rule It defines a rule output for one or more factor value combinations. See Rules (R-2.4).
Rule-Based Graph Graphical display of the hierarchy of attribute dependencies. See Rule-Based
Graph (R-2.3).
Rule-Based Knowledge Base Knowledge base, which uses Rule-Based Reasoning; therefore it contains
a Rule-Based Graph and rules. See Tender (R-7.1.1).
Rule-Based Reasoning (RBR) The process of determining case features of dependent and decision
attributes on basis of case features of input attributes using the rules. See Cases Deductive Reasoning
(R-2.2.4.1).
Rule domain The factor value combinations on which the rule is valid (it is active if it is not hidden).
Rule output A value that the rule assigns to the rule domain. See Rules (R-2.4).
Value range A range of values of one factor for which the rule is valid. See Rules Rule List & Rule
Processing (R-2.4.1.2).
Rule set Set of rules, usually considered for one node of the Rule-Based Graph. See Rules Rule
List & Rule Processing (R-2.4.1.2).
Unknown A case feature or rule output that has none of the values of the attribute.
Value A grade of satisfying an attribute. See Attributes (R-2.1).
R-95

Doctus Documentation

Uploaded by

Copyright:

Available Formats

Doctus Documentation

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

Doctus Documentation

Uploaded by

Copyright:

Available Formats

Doctus Documentation

In the seventies were great prospects in the applications of Artificial

2 About This Document

Maths: Logical and/or mathematical interpretations of features of Doctus, which may be

Technical: Details of technical solutions of Doctus, excluded from regular explanations.

3 List of Figures and Tables

3.1 List of Figures

Figure G-1 Building a Knowledge Base

Figure R-49 Choosing running mode

3.2 List of Tables

Table R-1 Keyboard Shortcuts

Figure G-1: Building a Knowledge Base

2 Original Decision (when there is no experience in the domain)

2.2 Acquisition of Attributes

Figure G-2: Acquiring Attributes and Values.

2.3 Hierarchy of the Attributes: the Rule-Based Graph

Figure G-3: The Rule-Based Graph.

2.4 Acquisition of Cases

Figure G-4: Acquisition of Cases.

2.5 Knowledge Import Data Mining

Figure G-5: Knowledge Import.

Figure G-6: Clustering External Data.

2.6 The Rules

If a rule connects one value of each factor, it is called elementary rule.

Maths: Use the markings: Attributes: A, B, C, (X is the decision attribute)

Rules: A=a1 B=b2 C=c1 X=x1

Read: If A is a1 and B is b2 and C is c1 and then X is x1

Maths: Use the markings from above.

Rule: A[a2, a5] B=b2 C=c1 X=x1

Read: If A is between a1 and a5 and B is b2 and C is c1 and then X is x1

Rule: A a2 B=b2 C=c1 X=x1

Read: If A is better or equal to a2 and B is b2 and C is c1 and then X is x1

Figure G-7: Rules in 1D.

Figure G-8: Rules in 2D.

Figure G-9: Advices for Input of Rules.

Figure G-10: Checking the Consistency of the Rule Set.

2.7 The Reasoning

Figure G-11: Results of Rule-Based Reasoning.

2.8 Decision Analyses and Fine-Tuning

Figure G-12: Explaining the Results.

Fine-tuning usually implies one or more of the followings:

deleting existing attributes and/or values

2.9 Knowledge Export Intelligent Portal

Figure G-13: Knowledge Export.

The exported knowledge may be:

Figure G-14: Evaluation in Intelligent Portal.

3.2 Acquisition of Attributes

Figure G-15: Acquiring Attributes and Values.

3.3 Acquisition of Cases

Figure G-16: Acquisition of Cases.

3.4 Knowledge Import Data Mining

Figure G-17: Knowledge Import.

Figure G-18: Clustering External Data.

3.5 Decision Tree: the Case-Based Graph

Figure G-19: The Case-Based Graph.

3.6 Classification of Cases

Maths: Determining informativity can be found in Appendix Determining

3.7 The Reasoning

3.8 Decision Analyses and Fine-Tuning

Figure G-20: Informativity and Density of the Attributes.