EXPERT SYSTEMS

1
EXPERT SYSTEMS

• CONTENTS
• 1. History of Expert Systems
• 2. Basic Concepts of Expert Systems
• 3. Knowledge Representations
• 4. Basic Architecture of an Expert System
• 5. Other Architectures of Expert Systems
• 6. Reasoning with Uncertainty
• 7. Tools for Building Expert Systems
• 8. Application Areas of Expert Systems
• 9. Benefits, Problems and Limitations of Expert
  Systems

2

• Example.
• General Electric's (GE)
• Top locomotive field service engineer was
  nearing retirement
• Traditional Solution apprenticeship
• Good short-term solution
• But GE wanted
• A more effective and dependable way to
  disseminate expertise.
• To prevent valuable knowledge from retiring.
• To minimize extensive travel or moving the
  locomotives.
• The expert system solution
• To model the way a human troubleshooter works
• Months of knowledge acquisition
• 3 years of prototyping
• A novice engineer or technician can perform at
  an experts level
• On a personal computer
• Installed at every railroad repair shop served
  by GE

3
History of Expert Systems
1. General Purpose Problem Solver
(mid-1960s) A. Newell and H.Simon few laws of
reasoning powerful computers
expertise performance initial situation
desired goal set of operators
4

• 2. Early Expert Systems (1970s)
• DENDRAL infers molecular structure from the
  unknown compounds
• MYCIN medical diagnosing (bacterial infections
  of the blood)
• E.Feigenbaum (Stanford University)
• "The expert knowledge provides the key to expert
  performance, while the knowledge representation
  and inference schemes provide the mechanism for
  its use."
• Conclusions
• General problem solvers are too weak.
• Human problem solvers are good only if they
  operate in a very narrow domain.
• Expert systems must be constantly updated with
  new information.
• The complexity of problems requires a
  considerable amount of knowledge about the
  problem area.

5
3. Development of Commercial Expert Systems
(1980) XCON - VAX System configuration (Digital
Equipment Corporation) YES/MVS - helps the
computer operator monitor and control the MVS
operating system (IBM) CATS-1 (Delta) - helps
diagnose and repair malfunctions in
diesel electric locomotives (General Electric
Company) Expert systems development
tools EMYCIN, EXPERT, META-DENDRAL, OPS5,
ROSIE, CRYSTAL, GOLDWORKS II, etc.
6
Basic Concepts of Expert Systems

• An expert system (ES) is a computer program that
  uses knowledgeand inference procedures of an
  expert to solve problems.
• Expertise is the extensive, task-specific
  knowledge acquired from training, reading, and
  experience.
• Includes
• theories about the problem area
• rules and procedures
• heuristics
• global strategies
• meta-knowledge
• facts
• Heuristics are rules of experience that
  characterise expert-level decision making in the
  field.

7

• Knowledge engineering - the task of eliciting and
  modelling the knowledge and building a computer
  system
• Transferring expertise
• expert computer non-expert
• knowledge acquisition (from experts and other
  sources)
• knowledge representation (in the computer)
• knowledge inferencing
• knowledge transfer to the user
• The human elements in expert systems
• expert
• knowledge engineer
• user

8

• shallow knowledge (experience)
• consists of all the peculiar heuristics and
  shortcuts
• e.g. IF it rains
• THEN the vegetables will grow faster
• deep knowledge (theoretical)
• first principles, axioms, laws
• e.g. IF it rains
• THEN the vegetables will grow faster
• BECAUSE the soil will become more moist
• Expert systems are tailored-made for specific
  and narrowly-defined problem domains.
• Second generation expert system employ both
• shallow deep knowledge

9

• Expert systems tasks
• interpretation - inferring situation
  descriptions from sensor data
• prediction - inferring likely consequences of
  given situations
• diagnosis - inferring malfunctions from
  observations
• prescription - prescribing remedies for
  malfunctions
• design - configuring objects under constraints
• planning - designing actions
• monitoring - comparing observations to expected
  outcomes
• control - governing overall system behaviour
• instruction - diagnosing, prescribing and
  guiding users' behaviour

10

• An ES has two main functions
• to draw conclusions
• to explain its reasoning
• Explanation
• HOW was a particular conclusion reached?
• WHY the program asks the user a particular
  question?
• TRACE displays all rules that are tried (not
  only those that are executed).
• WHAT-IF explains what will happen if a certain
  value or rule is changed.

11
Production Systems production condition-action
pair condition specifies data patterns action
specifies what to do example IF person A is
the mother of person B and person B is the mother
of person C THEN person A is the grandmother of
person C
12
example The Water Jug Problem two jugs 4
gallon, 3 gallon task to get exactly 2 gallons
of water in the 4 gallon jug State space (x, y)
where x 0, 1, 2, 3, 4 y0, 1, 2, 3 initial
state (0, 0) goal state (2, y) production rule
IF there are less than 4 gallons in the 4 gallon
jug THEN fill the 4 gallon jug IF (x, y) where x
lt 4 THEN (4, y)
13
(P1) IF (x, y) where x lt 4 THEN (4, y) (P2) IF
(x, y) where y lt 3 THEN (x, 3) (P3) IF (x, y)
where x gt 0 THEN (0, y) (P4) IF (x, y) where y gt
0 THEN (x, 0) (P5) IF (x, y) where xy gt 4 and y
gt 0 and x lt 4 THEN (4, y-(4-x)) (P6) IF (x, y)
where xy gt 3 and x gt 0 and y lt 3 THEN (x-(3-y),
3) (P7) IF (x, y) where xy ? 4 and y gt 0 THEN
(xy, 0) (P8) IF (x, y) where xy ? 3 and y gt 0
THEN (0, xy)
14

• The components of the production system
• (1) Working memory holds the symbol structure
  which represents the states of the problem
  space.
• (2) Production memory (long-term memory) holds
  the production rules.
• (3) Control strategy selects the rules to be
  applied.
• Conflict resolution
• principles
• ordering of the rules
• recency
• specificity

15

16
SUMMARY

• Expert systems imitate the reasoning process of
  experts.
• ES predecessor the General-purpose Problem
  Solver (GPS).
• Failed - ignored the importance of specific
  knowledge.
• The power of an ES is derived from its specific
  knowledge.
• Not from its particular knowledge representation
  or inferencescheme.
• Expertise is a task-specific knowledge acquired
  from training, reading, and experience.
• Experts can make fast and good decisions
  regarding complex situations.
• Expert system technology attempts to transfer
  knowledge from experts and documented sources to
  the computer and make it available to
  nonexperts.
• Expert systems provide limited explanation
  capabilities.