artificialintelligence-090417043101-phpapp02
TRANSCRIPT
-
7/29/2019 artificialintelligence-090417043101-phpapp02
1/69
1
Artificial Intelligence
A.I. can be define as the artificial brain having
capability of thinking and understanding.
A.I. is branch of computer science concerned with
the study and creation of computer system that exhibits
some form of intelligence.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
2/69
22/57
Knowledge-based systems
-
7/29/2019 artificialintelligence-090417043101-phpapp02
3/69
3
Introduction, or what is knowledge?
Knowledge Knowledge can be defined as the body of facts and
principles accumulated by human kind or the act ,fact or
state of knowing
is a theoretical or practical understanding of a subject The sum of what is currently known, and apparently
knowledge is power.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
4/69
4
In biological organisms, knowledge is likely stored as complex
structure of interconnected Neurons.
In computers, knowledge is stored as symbolic structure but in
form of collections of magnetic spots & voltage states
-
7/29/2019 artificialintelligence-090417043101-phpapp02
5/69
5
Knowledge SourcesDocumented (books, manuals, etc.)
Undocumented (in people's minds)
From people, from machines
Knowledge Acquisition from Databases
Knowledge Acquisition Via the Internet
-
7/29/2019 artificialintelligence-090417043101-phpapp02
6/69
6
Knowledge Levels
Shallow knowledge (surface) Deep knowledge
Can implement a computerized representation that is
deeper than shallow knowledge
Special knowledge representation methods (semanticnetworks and frames) to allow the implementation of
deeper-level reasoning (abstraction and analogy):
important expert activity
Represent objects and processes of the domain of
expertise at this level
Relationships among objects are important
-
7/29/2019 artificialintelligence-090417043101-phpapp02
7/69
7
Scope of Knowledge
Knowledge acquisition is the extraction of knowledge from
sources of expertise and its transfer to the knowledge base and
sometimes to the inference engine
Knowledge is a collection of specialized facts, procedures and
judgment rules
-
7/29/2019 artificialintelligence-090417043101-phpapp02
8/69
8
Domain Expert
Those who possess knowledge are called experts.
Anyone can be considered a domain expert if he or she hasdeep knowledge (of both facts and rules) and strong practicalexperience in a particular domain. The area of the domain maybe limited. In general, an expert is a skilful person who can do
things other people cannot. knowledgeable and skilled person capable of solving problems
in a specific area or domain.
Has the greatest expertise in a given domain.
This expertise is to be captured in the expert system.
Therefore, the expert must be able to communicate his orher knowledge, be willing to participate in the expert systemdevelopment and commit a substantial amount of time to theproject.
Most important player in the expert system developmentteam.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
9/69
9
Major Categories of Knowledge
Declarative Knowledge
Procedural Knowledge
Meta knowledge
-
7/29/2019 artificialintelligence-090417043101-phpapp02
10/69
10
Declarative Knowledge
Descriptive Representation of Knowledge
Expressed in a factual statement
Shallow
Important in the initial stage of knowledge acquisition
-
7/29/2019 artificialintelligence-090417043101-phpapp02
11/69
11
Procedural Knowledge
Considers the manner in which things work under different
sets of circumstances
Includes step-by-step sequences and how-to types of
instructions
May also include explanations
Involves automatic response to stimuli
May tell how to use declarative knowledge and how to
make inferences
-
7/29/2019 artificialintelligence-090417043101-phpapp02
12/69
12
Descriptive knowledge relates to a specific object.
Includes information about the meaning, roles,
environment, resources, activities, associations andoutcomes of the object
Procedural knowledge relates to the procedures employed
in the problem-solving process
-
7/29/2019 artificialintelligence-090417043101-phpapp02
13/69
13
Meta knowledge
Knowledge about Knowledge
Meta knowledge can be simply defined as knowledge aboutknowledge.
Meta knowledge is knowledge about the use and control ofdomain knowledge in an expert system.
In ES, Meta knowledge refers to knowledge about the operation ofknowledge-based systems
Its reasoning capabilities
-
7/29/2019 artificialintelligence-090417043101-phpapp02
14/69
14
Whats in the knowledge base?
Facts about the specifics of the world
Northwestern is a private university
The first thing I did at the party was talk to John.
Rules that describe ways to infer new facts from existing facts
All triangles have three sides
All elephants are grey
Facts and rules are stated in a formal languageGenerally some form of logic.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
15/69
15
Knowledge-based systems
A major turning point occurred in the field of AI with realization
that in knowledge lies the power.
This realization led to the development of a new class of
system: i.e. knowledgebased system.
knowledgebased system get their power from the expert
knowledge that has been coded into facts , rules & procedure.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
16/69
16
Components of KBS
The knowledge is stored in a knowledge base separated from
the control & inferencing component . This makes it possible
to add new knowledge or refine existing knowledge without
recompiling the control and inferencing programs.
Input output
unit
Inference control
unitKnowledge
base
-
7/29/2019 artificialintelligence-090417043101-phpapp02
17/69
1717/57
Structure and characteristics
AI programs:intelligent problem solving tools
KBSsAI programs with special programstructure separated knowledge base
ESs
KBSs applied in a specific narrow field
AI programs
Knowledge-based systems
Expert systems
-
7/29/2019 artificialintelligence-090417043101-phpapp02
18/69
18
The Knowledge Hierarchy
meta-
knowledge
knowledge
informationdata
noise
large volume, low
value, usually no
meaning/ context
lower volume, higher
value, with context and
associated meanings
understanding of a
domain, can be applied to
solve problems
knowledge on knowledge
(e.g how/when to apply)
may contain
irrelevant items
which obscure data
management
information
systems
knowledge-
basedsystems
databases,transaction
systems
-
7/29/2019 artificialintelligence-090417043101-phpapp02
19/69
19
Different type of knowledge base system
There are different type of knowledge base system as1. knowledge engineer
KE involves knowledge of applications of engineering field related to thatsystem.
2 Knowledge representation
The KR means the key topics & concept involved in the system berepresented using a data base system.
3 knowledge use
These are the utilization of knowledge for problem solving tech. thisinvolves analogical reasoning decision related methods and system etc.
4 Knowledge Acquisition
The knowledge has to be acquired by way of learning or by studying thesystem. The transfer of knowledge or exchange of knowledge results intoknowledge acquisition
`
-
7/29/2019 artificialintelligence-090417043101-phpapp02
20/69
20
What is Knowledge Engineering?
the process of building an ES
the effort in developing a large quantity of effective knowledge(i.e. the KB)
the acquisition of knowledge from a human expert or othersource (by a knowledge engineering) and its coding in the ES
KE is important, because:
performance of an ES is largely determined by the quantity &quality of knowledge in its KB
-
7/29/2019 artificialintelligence-090417043101-phpapp02
21/69
21
knowledge Engineering The process of building knowledge-based systems is called
knowledge engineering (KE). It has a great deal in common with
software engineering, and is related to many computer science
domains such as artificial intelligence, databases, data mining,
expert systems, decision support systems and geographicinformation systems. Knowledge engineering is also related to
mathematical logic and cognitive science as the knowledge is
produced by cognitive systems (mainly humans) and is
structured by our understanding of how human reasoning or
logic works.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
22/69
22
Knowledge Engineering
Art of bringing the principles and tools of AI research tobear on difficult applications problems requiring experts'
knowledge for their solutions
Technical issues of acquiring, representing and using
knowledge appropriately to construct and explain lines-of-reasoning
Art of building complex computer programs that represent
and reason with knowledge of the world(Feigenbaum and McCorduck [1983])
-
7/29/2019 artificialintelligence-090417043101-phpapp02
23/69
23
The knowledge engineer
someone who is capable of designing, building and testing
an expert system.
interviews the domain expert to find out how a particular
problem is solved.
establishes what reasoning methods the expert uses tohandle facts and rules and decides how to represent them
in the expert system.
chooses some development software or an expert system
shell, or looks at programming languages for encoding the
knowledge.
responsible for testing, revising and integrating the expert
system into the workplace.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
24/69
24
Knowledge Engineering
Process of acquiring knowledge from experts and building
knowledge base
Narrow perspective
Knowledge acquisition, representation, validation,
inference, maintenance
Broad perspective
Process of developing and maintaining intelligent system
-
7/29/2019 artificialintelligence-090417043101-phpapp02
25/69
25
Views of knowledge engineering
There are two main views to knowledge engineering:
Transfer View This is the traditional view. In this view,the assumption is to apply conventional knowledgeengineering techniques to transfer human knowledge into
artificial intelligence systems.
Modeling View This is the alternative view. In this view,the knowledge engineer attempts to model the knowledgeand problem solving techniques of the domain expert intothe artificial intelligence system.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
26/69
26
Knowledge Engineering Process Activities
Knowledge Acquisition
Knowledge Validation
Knowledge Representation
Inferencing
Explanation and Justification
-
7/29/2019 artificialintelligence-090417043101-phpapp02
27/69
27
Knowledge Engineering Process
Knowledge
validation
(test cases)
Knowledge
Representation
Knowledge
AcquisitionEncoding
Inferencing
Sources of knowledge
(experts, others)
Explanation
justification
Knowledge
base
-
7/29/2019 artificialintelligence-090417043101-phpapp02
28/69
28
Knowledge Engineering in a Nutshell
human
expert
knowledge
engineer
knowledge base
(in ES)
explicit
knowledge
dialog
knowledge
refinement
-
7/29/2019 artificialintelligence-090417043101-phpapp02
29/69
29
Phases of KE
Various phases of KE specific for the development of a knowledge-
based system:
* Assessment of the problem
* Acquisition and structuring of related information,knowledge and specific preferences
* Development of a knowledge-based system shell/structure
* Implementation of the structured knowledge into
knowledge-bases
* Testing and validation of the inserted knowledge* Integration and maintenance of the system
* Revision and evaluation of the system."
-
7/29/2019 artificialintelligence-090417043101-phpapp02
30/69
30
Knowledge Engineering Principles
Knowledge engineers acknowledge that there are different types of knowledge,and that the right approach and technique should be used for the knowledgerequired.
Knowledge engineers acknowledge that there are different types of expertsand expertise, such that methods should be chosen appropriately.
Knowledge engineers recognize that there are different ways of representingknowledge, which can aid the acquisition, validation and re-use of knowledge.
Knowledge engineers recognize that there are different ways of usingknowledge, so that the acquisition process can be guided by the project aims.
Knowledge engineers use structured methods to increase the efficiency of theacquisition process
-
7/29/2019 artificialintelligence-090417043101-phpapp02
31/69
31
The main players in the development team
Expert System
End-user
Knowledge Engineer ProgrammerDomain Expert
Project Manager
Expert System
Development Team
-
7/29/2019 artificialintelligence-090417043101-phpapp02
32/69
32
Intelligent SystemA System can be constructed as a intelligence system if it has fourmajor techniques of knowledge representation.
1.Logic
The logic is a formal procedure because of which implications arecreated from the set of known facts.
2.Production Systems
The production systems studies the new facts and the knownfacts and finds the desired conclusion.
3.Semantic networks
It is a network of symbols that describe relationship betweenelements of knowledge
4.Frames
These are the data structures which consists of expectations for agiven situation.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
33/69
33
Although knowledge representation is oneof the central and in some ways most
familiar concepts in AI, the most
fundamental question about itWhat is it?
has rarely been answered directly.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
34/69
34
What is a knowledge representation?
A knowledge representation (KR) is most fundamentally a surrogate, a substitute forthe thing itself, used to enable an entity to determine consequences by thinking rather
than acting, i.e., by reasoning about the world rather than taking action in it.
It is a set of ontological commitments, i.e., an answer to the question: In what termsshould I think about the world?
It is a fragmentary theory of intelligent reasoning, expressed in terms of threecomponents: (i) the representation's fundamental conception of intelligent reasoning;
(ii) the set of inferences the representation sanctions; and (iii) the set of inferences itrecommends.
It is a medium for pragmatically efficient computation, i.e., the computationalenvironment in which thinking is accomplished. One contribution to this pragmaticefficiency is supplied by the guidance a representation provides for organizinginformation so as to facilitate making the recommended inferences.
It is a medium of human expression, i.e., a language in which we say things about theworld.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
35/69
35
Elements of a Representation
Represented world: about what?
Representing world: using what?
Representing rules: how to map?
Process that uses the representation: conventions and systems
that use the representations resulting from above.
Analog versus Symbolic
-
7/29/2019 artificialintelligence-090417043101-phpapp02
36/69
36
Understanding the roles and acknowledging their diversity has several usefulconsequences. First, each role requires something slightly different from arepresentation; each accordingly leads to an interesting and different set ofproperties we want a representation to have.
Second, we believe the roles provide a framework useful for characterizing awide variety of representations. We suggest that the fundamental "mindset" ofa representation can be captured by understanding how it views each of theroles, and that doing so reveals essential similarities and differences.
Third, we believe that some previous disagreements about representation areusefully disentangled when all five roles are given appropriate consideration.We demonstrate this by revisiting and dissecting the early arguments
concerning frames and logic.
Finally, we believe that viewing representations in this way has consequencesfor both research and practice. For research, this view provides one directanswer to a question of fundamental significance in the field. It also suggestsadopting a broad perspective on what's important about a representation, andit makes the case that one significant part of the representation endeavor--capturing and representing the richness of the natural world--is receiving
insufficient attention.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
37/69
37
Terminology
Two points of terminology will assist in our presentation. First,we use the term inference in a generic sense, to mean anyway to get new expressions from old. We are only rarelytalking about sound logical inference and when doing so refer
to that explicitly.
Second, to give them a single collective name, we refer to thefamiliar set of basic representation tools like logic, rules,frames, semantic nets, etc., as knowledge representationtechnologies.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
38/69
38
We have argued that a knowledge representation plays five distinct roles,each important to the nature of representation and its basic tasks. Thoseroles create multiple, sometimes competing demands, requiring selectiveand intelligent tradeoff among the desired characteristics. Those five rolesalso aid in characterizing clearly the spirit of representations andrepresentation technologies that have been developed.
This view has consequences for both research and practice in the field. Onthe research front it argues for a conception of representation broader than
the one often used, urging that all of the five aspects are essentialrepresentation issues. It argues that the ontological commitment arepresentation supplies is one of its most significant contributions; hencethe commitment should be both substantial and carefully chosen. It alsosuggests that the fundamental task of representation is describing thenatural world and claims that the field would advance furthest by taking this
as its central preoccupation.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
39/69
39
Different levels of knowledge
representation
Mental Imag e
Written Text
Magnet ic Spots
Binary Numbers
Character String s
-
7/29/2019 artificialintelligence-090417043101-phpapp02
40/69
40
How Knowledge Representation Works
Intelligence requires knowledge
Computational models of intelligence require models ofknowledge
Use formalisms to write down knowledge
Expressive enough to capture human knowledge
Precise enough to be understood by machines
Separate knowledge from computational mechanisms thatprocess it
Important part of cognitive model is what the organismknows.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
41/69
41
How knowledge representations are
used in cognitive models
Contents of KB is
part of cognitive
model
Some models
hypothesizemultiple knowledge
bases.
Knowledge
Base
Inference
Mechanism(s)Learning
Mechanism(s)
Examples,Statements
Quest ions,requests
Answers,analyses
-
7/29/2019 artificialintelligence-090417043101-phpapp02
42/69
42
-
7/29/2019 artificialintelligence-090417043101-phpapp02
43/69
43
Knowledge acquisition
Knowledge acquisition includes the elicitation,
collection, analysis, modelling and validation ofknowledge forknowledge engineering and
knowledge management projects
http://www.epistemics.co.uk/Notes/61-0-0.htmhttp://www.epistemics.co.uk/Notes/40-0-0.htmhttp://www.epistemics.co.uk/Notes/40-0-0.htmhttp://www.epistemics.co.uk/Notes/61-0-0.htm -
7/29/2019 artificialintelligence-090417043101-phpapp02
44/69
44
-
7/29/2019 artificialintelligence-090417043101-phpapp02
45/69
45
Issues in Knowledge Acquisition
. Some of the most important issues in knowledge acquisition are asfollows:
Most knowledge is in the heads of experts
Experts have vast amounts of knowledge
Experts have a lot of tacit knowledge
They don't know all that they know and use
Tacit knowledge is hard (impossible) to describe
Experts are very busy and valuable people
Each expert doesn't know everything
Knowledge has a "shelf life"
-
7/29/2019 artificialintelligence-090417043101-phpapp02
46/69
46
Requirements for KA Techniques
Because of these issues, techniques are requiredwhich:
Take experts off the job for short time periods
Allow non-experts to understand the knowledge
Focus on the essential knowledge
Can capture tacit knowledge
Allow knowledge to be collated from different
expertsAllow knowledge to be validated and maintained
-
7/29/2019 artificialintelligence-090417043101-phpapp02
47/69
47
KA Techniques Many techniques have been developed to help elicit knowledge from an expert.
These are referred to as knowledge elicitation or knowledge acquisition (KA)techniques. The term "KA techniques" is commonly used.The following list
gives a brief introduction to the types of techniques used for acquiring,
analysing and modelling knowledge:
Protocol-generation techniques include various types of interviews
(unstructured, semi-structured and structured), reporting techniques (such asself-report and shadowing) and observational techniques
Protocol analysis techniques are used with transcripts of interviews or other
text-based information to identify various types of knowledge, such as goals,
decisions, relationships and attributes. This acts as a bridge between the use
of protocol-based techniques and knowledge modelling techniques.
Hierarchy-generation techniques, such as laddering, are used to build
taxonomies or other hierarchical structures such as goal trees and decision
networks.
KA Techniques
http://www.epistemics.co.uk/Notes/174-0-0.htmhttp://www.epistemics.co.uk/Notes/176-0-0.htmhttp://www.epistemics.co.uk/Notes/178-0-0.htmhttp://www.epistemics.co.uk/Notes/178-0-0.htmhttp://www.epistemics.co.uk/Notes/176-0-0.htmhttp://www.epistemics.co.uk/Notes/174-0-0.htmhttp://www.epistemics.co.uk/Notes/174-0-0.htmhttp://www.epistemics.co.uk/Notes/174-0-0.htm -
7/29/2019 artificialintelligence-090417043101-phpapp02
48/69
48
ec ques
Matrix-based techniques involve the construction of grids indicating such thingsas problems encountered against possible solutions. Important types include theuse of frames for representing the properties of concepts and the repertory gridtechnique used to elicit, rate, analyse and categorise the properties of concepts.
Sorting techniques are used for capturing the way people compare and orderconcepts, and can lead to the revelation of knowledge about classes, propertiesand priorities.
Limited-information and constrained-processing tasks are techniques that eitherlimit the time and/or information available to the expert when performing tasks.
For instance, the twenty-questions technique provides an efficient way ofaccessing the key information in a domain in a prioritised order.
Diagram-based techniques include the generation and use of concept maps, statetransition networks, event diagrams and process maps. The use of these isparticularly important in capturing the "what, how, when, who and why" of tasksand events.
http://www.epistemics.co.uk/Notes/182-0-0.htmhttp://www.epistemics.co.uk/Notes/184-0-0.htmhttp://www.epistemics.co.uk/Notes/184-0-0.htmhttp://www.epistemics.co.uk/Notes/180-0-0.htmhttp://www.epistemics.co.uk/Notes/186-0-0.htmhttp://www.epistemics.co.uk/Notes/188-0-0.htmhttp://www.epistemics.co.uk/Notes/188-0-0.htmhttp://www.epistemics.co.uk/Notes/188-0-0.htmhttp://www.epistemics.co.uk/Notes/188-0-0.htmhttp://www.epistemics.co.uk/Notes/186-0-0.htmhttp://www.epistemics.co.uk/Notes/186-0-0.htmhttp://www.epistemics.co.uk/Notes/186-0-0.htmhttp://www.epistemics.co.uk/Notes/186-0-0.htmhttp://www.epistemics.co.uk/Notes/186-0-0.htmhttp://www.epistemics.co.uk/Notes/180-0-0.htmhttp://www.epistemics.co.uk/Notes/184-0-0.htmhttp://www.epistemics.co.uk/Notes/184-0-0.htmhttp://www.epistemics.co.uk/Notes/182-0-0.htmhttp://www.epistemics.co.uk/Notes/182-0-0.htmhttp://www.epistemics.co.uk/Notes/182-0-0.htm -
7/29/2019 artificialintelligence-090417043101-phpapp02
49/69
49
Knowledge Acquisition Methods:
An Overview
Manual
Semiautomatic
Automatic (Computer Aided)
Manual Methods Structured
-
7/29/2019 artificialintelligence-090417043101-phpapp02
50/69
50
Manual Methods Structured
Around
Interviews
Process
Interviewing
Tracking the Reasoning Process
Observing
Manual methods: slow, expensive and sometimes
inaccurate
-
7/29/2019 artificialintelligence-090417043101-phpapp02
51/69
51
Manual Methods of knowledge Acquisition
Knowledge
base
Documented
knowledge
Experts
CodingKnowledge
engineer
-
7/29/2019 artificialintelligence-090417043101-phpapp02
52/69
52
Semiautomatic Methods
Support Experts Directly
Help Knowledge Engineers
-
7/29/2019 artificialintelligence-090417043101-phpapp02
53/69
53
Expert-Driven
Knowledge Acquisition
Knowledge
base
Knowledge
engineer
ExpertCodingComputer-aided
(interactive)
interviewing
-
7/29/2019 artificialintelligence-090417043101-phpapp02
54/69
54
Automatic Methods
Experts and/or the knowledge engineers roles are
minimized (or eliminated)
Induction Method.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
55/69
55
Induction-Driven Knowledge Acquisition
Knowledge
base
Case histories
and examples
Induction
system
-
7/29/2019 artificialintelligence-090417043101-phpapp02
56/69
56
-
7/29/2019 artificialintelligence-090417043101-phpapp02
57/69
57
-
7/29/2019 artificialintelligence-090417043101-phpapp02
58/69
58
-
7/29/2019 artificialintelligence-090417043101-phpapp02
59/69
59
-
7/29/2019 artificialintelligence-090417043101-phpapp02
60/69
60
Knowledge Acquisition Difficulties
Problems in Transferring Knowledge
Expressing Knowledge
Transfer to a Machine
Number of Participants
Structuring Knowledge
Other Reasons
-
7/29/2019 artificialintelligence-090417043101-phpapp02
61/69
61
Experts may lack time or not cooperate
Testing and refining knowledge is complicated
Poorly defined methods for knowledge elicitation
System builders may collect knowledge from one source, but the
relevantknowledge may be scattered across several sources
May collect documented knowledge rather than use experts
The knowledge collected may be incomplete
Difficult to recognize specific knowledge when mixed with irrelevant
data
Other Reasons
-
7/29/2019 artificialintelligence-090417043101-phpapp02
62/69
62
Experts may change their behavior when observed and/orinterviewed
Problematic interpersonal communication between theknowledge engineer and the expert
Critical
The ability and personality of the knowledge engineer
Must develop a positive relationship with the expert
The knowledge engineer must create the right impression
Computer-aided knowledge acquisition tools
Extensive integration of the acquisition efforts
-
7/29/2019 artificialintelligence-090417043101-phpapp02
63/69
6363/57
Advantages of KBSs and ESs
make up for shortage of experts, spread expert knowledge on availableprice
field of interest changes are well-tracked
increase expert ability and efficiency
preserve know-how
can be developed systems unrealizabled with tradicional technology(Buck Rogers)
self-consistents in advising, equable in performance are availablepermanently
able to work even with partial, non-complete data
able to give expanation
-
7/29/2019 artificialintelligence-090417043101-phpapp02
64/69
6464/57
Disadvantages of KBSs and ESs
their knowledge is from a narrow field, dont know the limits
the answers are not always correct (advices have to beanalysed!)
dont have common sence (greatest restriction) all of theself-evident checking have to be defined
(many exceptions increase the size of KB and the runningtime)
Role V: A KR is a Medium of Human Expression
-
7/29/2019 artificialintelligence-090417043101-phpapp02
65/69
65
Finally, knowledge representations are also the means bywhich we express things about the world, the medium ofexpression and communication in which we tell the machine(and perhaps one another) about the world. This role forrepresentations is inevitable so long as we need to tell themachine (or other people) about the world, and so long as wedo so by creating and communicating representations. (5) Thefifth role for knowledge representations is thus as a medium ofexpression and communication for use by us.
http://groups.csail.mit.edu/medg/ftp/psz/k-rep.htmlhttp://groups.csail.mit.edu/medg/ftp/psz/k-rep.html -
7/29/2019 artificialintelligence-090417043101-phpapp02
66/69
66
Role IV: A KR is a Medium forEfficient Computation
From a purely mechanistic view, reasoning inmachines (and somewhat more debatably, in people)
is a computational process. Simply put, to use arepresentation we must compute with it. As a result,questions about computational efficiency are inevitablycentral to the notion of representation.
Role III: A KR is a Fragmentary Theory Of Intelligent Reasoning
-
7/29/2019 artificialintelligence-090417043101-phpapp02
67/69
67
The third role for a representation is as a fragmentary theory ofintelligent reasoning. This role comes about because the initialconception of a representation is typically motivated by someinsight indicating how people reason intelligently, or by somebelief about what it means to reason intelligently at all.
The theory is fragmentary in two distinct senses: (i) therepresentation typically incorporates only part of the insight or
belief that motivated it, and (ii) that insight or belief is in turn onlya part of the complex and multi-faceted phenomenon ofintelligent reasoning.
A representation's theory of intelligent reasoning is often implicit,but can be made more evident by examining its threecomponents: (i) the representation's fundamental conception ofintelligent inference; (ii) the set of inferences the representationsanctions; and (iii) the set of inferences it recommends.
-
7/29/2019 artificialintelligence-090417043101-phpapp02
68/69
68
Role II: A KR is a Set of Ontological Commitments
If, as we have argued, all representations are imperfect
approximations to reality, each approximation attending tosome things and ignoring others, then in selecting any
representation we are in the very same act unavoidably making
a set of decisions about how and what to see in the world. That
is, selecting a representation means making a set of
ontological commitments. (2) The commitments are in effect astrong pair of glasses that determine what we can see, bringing
some part of the world into sharp focus, at the expense of
blurring other parts.
Role I: A KR is a Surrogate
http://groups.csail.mit.edu/medg/ftp/psz/k-rep.htmlhttp://groups.csail.mit.edu/medg/ftp/psz/k-rep.html -
7/29/2019 artificialintelligence-090417043101-phpapp02
69/69
Role I: A KR is a Surrogate
Any intelligent entity that wishes to reason about its worldencounters an important, inescapable fact: reasoning is aprocess that goes on internally, while most things it wishes toreason about exist only externally. A program (or person)engaged in planning the assembly of a bicycle, for instance,may have to reason about entities like wheels, chains,sprockets, handle bars, etc., yet such things exist only in theexternal world.
This unavoidable dichotomy is a fundamental rationale and rolefor a representation: it functions as a surrogate inside thereasoner, a stand-in for the things that exist in the world.Operations on and with representations substitute for operations
on the real thing, i.e., substitute for direct interaction with theworld. In this view reasoning itself is in part a surrogate foraction in the world, when we can not or do not (yet) want to takethat action. (1)
http://groups.csail.mit.edu/medg/ftp/psz/k-rep.htmlhttp://groups.csail.mit.edu/medg/ftp/psz/k-rep.html