2 of 50

UNIT-5: Expert Systems

Architecture of expert systems, Roles of expert systems Knowledge Acquisition Meta knowledge Heuristics. Typical expert systems - MYCIN, DART, XCON: Expert systems shells.

3 of 50

Artificial Intelligence

The ability of computers to duplicate the functions of the human brain

4 of 50

Interesting Statistics

It has been estimated that computers that can exhibit humanlike intelligence (including musical and artistic aptitude, creativity, physical movement physically, and emotional responsiveness) require processing power of 20 million billion calculations per second (by the year 2030?).

5 of 50

The Difference Between Natural & Artificial Intelligence


Use Sensors	High	Low
Creativity and Imagination	High	Low
Learn from Experience	High	Low
Adaptability	High	Low
Access external information	High	Low
Make complex calculations	Low	High
Transfer information	Low	High

6 of 50

The Major Branches of AI(application of AI)

Viîon Systems

Natural language

Processing

7 of 50

Expert Systems (ES)

What are Expert Systems?

he expert systems are the computer applications developed to solve complex problems in a

particular domain, at the level of extra-ordinary human intelligence and expertise.

An expert system compared wlth tradltlonal computer :

Inference engine Knowledge = Expert system

( Algorithm -I- Data structures = Program in traditional computer )

First expert system, called DENDRAL, was developed in the early 70's at

Stanford University.

8 of 50

Characteristics of Expert Systems

High performance
Understandable
Reliable
Highly responsive

9 of 50

Capabilities of Expert System

he expert systems are capable of —

Advising
instructing and assisting human in decision making Demonstrating
Deriving a solution

Interpreting input
Predicting results
justifying the conclusion
Suggesting alternative options to a problem

hey are incapable of —

Substituting human decision makers

Possessing human capabilities

Producing accurate output for inadequate knowledge base
Refining their own knowledge

10 of 50

Components of Expert System

Components of Expert Systems

The components of ES include —

Knowledge Base
Interface Engine
User Interface

11 of 50

Components of ES

Domain

Expert

Expertise

User

Knowledge Engineer

Encoded

Expertise

Components of Expert System

System

Engineer

12 of 50

Component of an expert system

User

Expert System shell

Inter

face

Explanation System

Inference ence

Knowledge

base editor

Case specific data: Worlfing storage

Knowledge base

13 of 50

The Knowledge Base

Stores all relevant information, data, rules, cases, and

relationships used by the expert system

^ Assembling human experts

Use of fuzzy logic

A special research area in computer science that allows shades of gray and does not require everything to be simple black/white, yes/no, or true/false

^ Use of rules

Conditional statement that links given conditions to actions

or outcomes

E.g. if-then statements

* Use of cases

14 of 50

Components of Knowledge Base

*The knowledge Dase of an ES is a store of both, factual and heuristic knowledge.

Factual Knowledge — It is the information widely accepted by the Knowledge Engineers and scholars in the task domain.

Heuristic Knowledge — It is about practice, accurate judgement, one's a bility of evaluation,

and guessing.

Knowledge representation

It is the method used to organize and formalize the knowledqe in the knowledge base. It is in the form of IT-THEN-ELSE ru es.

Knowledge Acquisition

The success of any expert system majorly depends on the quality, completeness, and accuracy of the information stored in the knowledge base.

The knowledge base is formed by readings from various experts, scholars, and the Knowledge Engineers. The knowledge engineer is a person with the qualities of empathy, quick learning, and case analyzinq skills.

He and observin him at

work.

15 of 50

The Inference Engine

Figure : Rules for a Credit Application

Nortgaga Application for Loans money Ct00,000 to

$200.0iX/

If there are no previous credit problems and

If monthly net income is prior than

If down payment is t5% of the total value of the property and

If net assets of borrower are greater than

$25.000 and

If employment is greater than three years at

the

Then accept loan application

Else check ether credit rules

16 of 50

To recommend a solution, the interface engine uses the following strategies —

Forward Chaining
Backward Chaining

17 of 50

Forward Chaining

It is a strategy of an expert system to answer the question,"What can happen naxt?"

Here, the interface engine follows the chain of conditions and derivations and finally deduces the

outcome. It considers all be facts and ales, and sorb them before concluding to a solution.

This strategy is followed for working on conclusion, result, or effect. For example, prediction of share market status as an effect of changes in interest rates.

Fact 1

AND

Fact3

Fact4

18 of 50

Backward Chaining

JitL this strategy, ar expert system finds out the answer to the question, “Why this happened?"

3n the basis of what has already Lappered, the interface engine tries to find out which conditions ^c0UId^have^tappered^ir^the^past^f0r^this^resut^.^ThiS^stratepy^iS^{f0IIoweü for finding}^out^cavse^orreas0r For example, diagnosis of bl cd cancer in humans.

19 of 50

Expert system Technology

Expert System Development Environment — The ES development environment includes haroware and tools. They are —

Workstations, minicomputers, mainframes.
High level Symbolic Programming Languages such as LlSt Pro PROgrammation en LOGique PROCOG.

Large databases.

Tools - They reduce the effort and cost involved in developing an expert system to large extent.

Powerful editors and debu tools with multi-windows.

They provide rapid prototyping
Have inbuilt definibons of model, knowledge representation, and inference design.

Shells — A shell is nothing but an ex ert systemwithout knowledge base. A shell provides the developers with knowledge acquisition, inference engine, user interface, and explanation facility. For example, few shells are given below -

lava Expert System Shell JESSthat provides fully developed java API for creating anexpert system.
Vidwan, a shell developed at the National Centre for Software Technology, Mumbai in

1993. It enables knowledge encoding in the form of IF-THEN rules.

20 of 50

Expert Systems Development

Determining requirements

Identifying experts

Constructing expert

Implementing results

components

Figure : Steps in the Expert System Development Process

and reviewing system

21 of 50

Participants in Expert System

Development

Domain

The area of knowledge addressed by the expert

system

Domain Expert

The individual or group who has the expertise or knowledge one is trying to capture in the expert system

Knowledge Engineer

An individual who has training or expertise in the design, development, implementation, and maintenance of an expert system

Knowledge User

The individual or group who uses and benefits from the

expert system

22 of 50

Application of ES

Design Domain

Pledical Domain

Monitoring Systems

Process Control Systems Knowledge Domaln Finance/Commerce

Description

Camera lens design, automobile design.

Diagnosis Systems to deduce cause of disease from observed data, conduction medical operabons on humans.

Comparing data continuously wlth observed system or with prescribed behavior such as leakage monitoring in long petroleum pipeline.

Controlling a physical process based on monitoring. Finding out faulb in vehicles, computers.

Detection of possible fraud, suspicious transactions, stock market trading, Airline scheduling, cargo scheduling.

23 of 50

Benefits of Expert System

Availability – They are easily due to mass production of software.
Less Production Cost – Production cost is reasonable. This makes them affordable

Speed– They offer great speed. They reduce the amount of work an individuals puts in.
Reducing Risks – They can work in the environments dangerous to humans.

24 of 50

Limitations of an Expert System

Not widely used or tested
Difficult to use
Limited to relatively narrow problems
Possibility of error
Cannot refine its own knowledge
Difficult to maintain

25 of 50

Expert System Shells

^ In the 1980s, expert system "shells were introduced and supported the development o1 expert systems in a

of LISP code was

wide variety of application areas.

^ During the work ,a large amount

written for different modules:

Knowledge base
Inference engine

Working memory
Explanation facility
End-user interface

26 of 50

Expert System Shells

An Expert system shell is a software development environment. It contains the basic components of expert systems. A shell is associated with a prescribed method for building applications by configuring and

instantiating these components.

Shell components and description

The generic components of a shell :

knowledge Base, the reasoning,

the knowledge

the explanation

acquisition, the

and the user

interface are

are the core

shown below. The

components.

knowledge base and reasoning engine

29 of 50

MYCIN was an early expert system that used artificial intelligence to identify bacteria causing severe infections.

^ recommend antibiotics, with the dosage adjusted for patient's body weight

+ The MYCIN system was also used for the diagnosis of blood clotting diseases.

MYCIN was developed over five or six years in the early 1970s at Stanford University.

+ It was written in Lisp

30 of 50

MYCIN was a standalone system that required a user to enter all relevant information about a patient by typing in responses to questions MYCIN posed.
MYCIN operated using a fairly simple inference engine, and a knowledge base of ~600 rules.

* It would query the physician running the program via

a long series of simple yes/no or textual questions.

31 of 50

Tasks and Domain

^ Disease DIAGNOSIS and Therapy

SELECTION

^ Advice for non-expert physicians with time

considerations and incomplete evidence on:

Bacterial infections of the blood
Expanded to meningitis and other ailments
Meet time constraints of the medical field

32 of 50

MYCIN Architecture

Consultation

Knowledge Acquisition System

Rułes

Tables, Lłsts

33 of 50

Consultation System

Performs Diagnosis and Therapy Selection
Control Structure reads Static DB (rules) and read/writes to Dynamic DB (patient, context)
Linked to Explanations
Terminal interface to Physician

34 of 50

Consultation “Control

Structure”

Goal-directed Backward-chaining Depth-first

Tree Search

High-level Algorithm:

Determine if Patient has significant infection
Determine likely identity of significant organisms
Decide which drugs are potentially useful

4. Select best drug or coverage of drugs

35 of 50

tatic Database

Rules
Meta-Rules
Templates

^ Rule Properties

Context Propei1ies
Fed from Knowledge Acquisition System

36 of 50

ynamic Database

Patient Data
Laboratory Data
Context Tree

+ Built by Consultation System

Used by Explanation System

37 of 50

xplanation System

+ Provides reasoning why a conclusion has been made, or why a question is being asked

Q-A Module
Reasoning Status

Checker

38 of 50

DART

DART is a joint project of the Heuristic Programming Project and IBM that explores the application of artificial intelligence techniques to the diagnosis of computer faults.

The primary goal of the DART Project is to develop programs that capture the special design knowledge and diagnostic abilities of these experts and to make them available to field engineers.

The practical goal

an automated

the construction

of pinpointing the

units

diagnostician capable responsible for

observed malfunctions in

functional arbitrary system

configurations.

39 of 50

Dynamic Analysis and Replanning Tool

DART uses intelligent agents to aid decision support system Give planners the ability to rapidly evaluate plans for logistical feasibility.
DART decreases the cost and time required to implement

decisions.

The field engineer is familiar with the diagnostic equipment and software testing.

Access to information about the specific system hardware and software configuration of the installation.

40 of 50

Xcon

The R1 (internally called XCON, for eXpert CONfigurer) program was a production rule based system written in OPS5 by John P. McDermott of CMU in 1978.

configuration of DEC VAX computer systems

ordering of DEC's VAX computer systems by automatically selecting the computer system components based on the customer's requirements.

XCON first went into use in 1980 in DEC's plant in Salem, New Hampshife. It eventually had about 2500 rules.

By 1986, it had processed 80,000 orders, and achieved 9598% accuracy.

It was estimated to be saving DEC $25M a year by reducing the need to give customers free components when technicians made errors, by speeding the assembly process, and by increasing customer satisfaction.

41 of 50

XCON interacted with the sales person, asking critical questions before

printing out a coherent and workable system specification/order slip.

XCON's success led DEC to rewrite XCON as XSELa version of XCON intended for use by DEC's salesforce to aid a customer in properly configuring their VAX.

42 of 50

XCON: Expert Configurer

Stages of Expert System building

< Identif cation:

Problems, data, goals, company, people...

Conceptualization:

Characterize different kinds of concepts and relations

Forma ization:

Express character of search

Imp ementation:

Build the system in executable form

Test ng and Eva uat

Does it do what we wanted?

Maintenance

Adapt to changing environment or requirements

43 of 50

Phase 1: Identification

DEC, Digital Equipment Corporation

Large computer manufacturer, started 1957

Catalogue has 40.000 different parts

Buyer (with Sales Rep) sends order, typically 100 parts

Delivery and assembly by DEC personnel

Too often, part collection does not allow installation

Too often, installed computer does not meet requirements Remedy: Completely assemble and test system in factory Automate configuration problem;

attempts with procedural languages were unsuccessful

XS approach started around 1980

44 of 50

Phase 2: Conceptualization

45 of 50

Phase 3: Formalization

Configuration engineers could talk well to Knowledge Engineers of the CSDG
Could explain in what stage which component should be configured how

^ This was expressed in production rules IF c1, c2 c3 THEN a1, a2, a3

+ Configuration stage was explicitly represented as data: current goal or context

Changing contexts moved configuration process through all stages

46 of 50

Phase 4: Implementation into system R1

Language: OPS5 (similar to CLIPS)
Conflict Resolution: MEA (extends Lex / Specificity)
Means-Ends Analysis: order by decency of first condition IF c1, c2 THEN .. is now different from IF c2, c1 THEN
Contexts are treated as special by putting them first
End-task is unspecific, thus executed last
Use MEA + Spec to concentrate on subtasks:

IF g1, x, y THEN assert barify II Signal necessity of subtask

IF barify, a THEN p, q
IF barify, b THEN r, s

II Two rules perform the task

II of barification per se

IF barify ready

THEN retract barify // Termination when

47 of 50

Field test after 1 year, production after another year
Accuracy over 95% :

No more pre-assembly nos necessary!

The installed configurations are optimized:

Buyers are happier because they see better products

Less retraining of staff on product changes:

Quicker change of production

Net return to Dig itaI is estimated to 40M $ per year.

But this name...

Field name: XCON

Phase 5: Quality and Testing of R1

48 of 50

Users of XCON

Key Roles

Sales: Use for quotations

Cha mp ion : Vision, belief in technology, influence on sponsor.
Sponsor: Has interest in problem and control over resources: money 1 peop Ie
Nlanager: Keeps parties

together, goals realistic

Knowledge E ng ineer: AI and