Part 07

Expert Systems

Third Year Information Technology

MANAGEMENT INFORMATION SYSTEM

Tushar B Kute,
Department of Information Technology,
Sandip Institute of Technology and Research Centre, Nashik
http://www.tusharkute.com
EXPERT SYSTEM ARCHITECTURE (1)
 The typical architecture of an e.s. is often
described as follows:

user inference knowledge

user
engine base
interface
EXPERT SYSTEM ARCHITECTURE (1)

 The inference engine and knowledge base

are separated because:
 the reasoning mechanism needs to be as stable
as possible;
 the knowledge base must be able to grow and
change, as knowledge is added;
 this arrangement enables the system to be built
from, or converted to, a shell.
EXPERT SYSTEM ARCHITECTURE (2)

 It is reasonable to produce a richer, more

elaborate, description of the typical expert
system.
 A more elaborate description, which still
includes the components that are to be found
in almost any real-world system, would look
like this:
EXPERT SYSTEM ARCHITECTURE (2)
EXPERT SYSTEM ARCHITECTURE (2)
EXPERT SYSTEM ARCHITECTURE (2)
 The system holds a collection of general
principles which can potentially be applied to
any problem - these are stored in the knowledge
base.
 The system also holds a collection of specific
details that apply to the current problem
(including details of how the current reasoning
process is progressing) - these are held in
working memory.
 Both these sorts of information are processed by
the inference engine.
EXPERT SYSTEM ARCHITECTURE (2)

 Any practical expert system needs an

explanatory facility. It is essential that an
expert system should be able to explain its
reasoning.
 EXPERT & KNOWLEDGE-BASED SYSTEMS
 One of AI’s greatest areas of success was the development
of large-scale problem solving systems
 Originally called expert systems, they would mimic the problem
solving processes of domain experts
 Such as doctors performing diagnosis, or engineers performing design,
or wall street analysts selecting stock transactions
 Expert systems were originally developed by hand
 And most commonly in some Lisp dialect
 It was discovered that many problems were being solved by
chaining through rules (if-then statements) that would
operate on a collection of facts and partial conclusions
 Called working memory
 These rule-based systems led to the first AI tools or shells
 Today, to simplify expert system creation, most people use these AI
shells – you just fill in the knowledge, the problem solving processes
are already implemented
 INTRODUCTION: DENDRAL
 The Dendral system (DENDRitic ALgorithm) was
the first expert system, developed in the 1960s
 The idea was, given mass spectrogram data, determine
what the chemical composition was
 The approach: plan-generate-and-test with human
feedback
 This is a constrained search technique
 Generate a hypothesis: a possible chemical compound
 Test the hypothesis: use a series of heuristics and subprograms
to determine if the chemical compound generated is plausible
given the data
 If so, show it to the user – the user can steer the direction that
DENDRAL takes next by suggesting what chemical elements
should be added or removed from the generated hypothesis
compound
 DENDRAL repeats until an answer is found acceptable that does
not violate the constraints as dictated by the data
 PATTERN MATCHING
 Expert Systems consisted of two major components
 A knowledge base
 The knowledge that the system would use to come to conclusions
 In Dendral, this would be chemical elements, compounds, the mass
spectrogram data that could arise from any element
 In Mycin, this would be bacteriological infection knowledge
 An inference engine
 How to apply the knowledge
 For Dendral, this was done by generate-and-test, a very weak form of search
 Mycin would perform backward chaining through rules
 More generically though, the inference engine would embody some
form of pattern matching
 One of the first systems that clearly demonstrated pattern matching
was Eliza
 Eliza was a spoof of an expert system but the code will show us how
pattern matching works
 ELIZA
 Created by Joseph
 Sample dialog:
Weizenbaum
 H: Men are all alike
 This system would act like a
 E: In what way?
Rogerian psychoanalyst by
 H: They’re always bugging us about
holding a dialog with a
something or another.
person
 E: Can you be more specific?
 the dialog would be consist
of the doctor (Eliza) asking  H: He says I’m depressed much of the
questions, the human time.
responding, and the doctor  E: I am sorry to hear you are depressed.
using the response to ask  H: It’s true. I am unhappy.
another question  E: Do you think coming here will help you
 Weizenbaum wrote the not to be unhappy?
program so that it would  H: I need some help, that much seems
generate an English certain.
response/question based on  E: What would it mean to you if you got
a group of patterns some help?
 If the user sentence  H: Perhaps I could learn to get along with
matched a pattern, this my mother.
pattern would be used to  E: Tell me more about your family
generate the next
 HOW ELIZA WORKS
 Repeat
 Input a sentence
 Find a rule in the Eliza knowledge-base that matches the
pattern
 Attempt to perform pattern match (see next slide)
 Attempt to perform segment match (see two slides)
 If rule found, select one of the responses randomly (each
pattern will have at least one response)
 Fill in any variables
 Substitute values (you for I, I for you, me for you, am for
are, etc)
 Respond
 Until user quits
 Most of the work goes on in the pattern match and
segment match functions
ARTIFICIAL INTELLIGENCE SYSTEMS
 It falls into three basic categories:
 Expert systems.
 Natural Language Processing
 Perception Systems.
 AI is a software technique applied to the non-
numeric data expressed in terms of symbols,
statements and patterns.
 It uses methods of symbolic processing, social and
scientific reasoning and conceptual modeling for
solving the problems.
CATEGORIES OF AI

Artificial Intelligence System

Natural
Expert Perception
Language

Uses Uses Uses

Native Language Size, Shape,
Knowledge
Knowledge Image, Voice

Applies Applies Applies

Language Human Like Sensing Abilities
Reasoning Reasoning for Reasoning
 AI Applications

Uses Human Uses Computer Uses Human

Information Intelligence for capabilities in
Processing producing speech
Capability Human Like recognition,
Capacity Multi Sensory
Interfacing
AI Applications Robotics
Applications Natural Interface
Applications
Intelligent Agents
Fuzzy Logic Robot Systems
Learning System for doing Human
Expert System Jobs VR Systems
KNOWLEDGE BASED EXPERT SYSTEMS
 Decision making or problem solving is a unique
situation riddled with uncertainty and complexity,
dominated by resource constraints and a possibility
of several goals. In such cases, flexible systems
(open systems) are required to solve the problems.
 Most of such situations, termed as the unstructured
situations, adopt two methods of problem solving,
generalized or the knowledge based expert
systems.
KBES
 To build a KBES, certain prerequisites are required.
The first prerequisite is that a person with the ability to
solve the problem with knowledge based reasoning
should be available.
 Second prerequisite is that, such an expert should be
able to articulate the knowledge to the specific problem
characteristics.
 Knowledge in KBES is defined as a mix of theory of the
subject, knowledge of its application, organized
information and the data of problems and its solutions.
REFERENCE

 Waman Jawadekar, "Management Information

Systems” , 4th Edition, Tata McGraw-Hill
Publishing Company Limited.
 E. Turban, J. Aronson, T.P. Liang, R. Sharda,
“Decision Support and Business Intelligence
Systems”, 8th Edition, Pearson Education.