Chapter 3

Solving Problems by Searching

Reflex agent is simple
 base their actions on
 a direct mapping from states to actions

 but cannot work well in environments

 whichthis mapping would be too large to store
 and would take too long to learn

Hence, goal-based agent is used

 Problem-solving agent
Problem-solving agent
A kind of goal-based agent
 It solves problem by
 finding sequences of actions that lead to
desirable states (goals)
 To solve a problem,
 the first step is the goal formulation, based on
the current situation
 Goal formulation
The goal is formulated
 as a set of world states, in which the goal is
satisfied
Reaching from initial state  goal state
 Actions are required
Actions are the operators
 causing transitions between world states
 Actions should be abstract enough at a
certain degree, instead of very detailed
 E.g., turn left VS turn left 30 degree,
degree etc.
 Problem formulation
The process of deciding
 what actions and states to consider
E.g., driving Amman  Zarqa
 in-between states and actions defined
 States: Some places in Amman & Zarqa

 Actions: Turn left, Turn right, go straight,

accelerate & brake, etc.

 Search
Because there are many ways to achieve
the same goal
 Those ways are together expressed as a tree
 Multiple options of unknown value at a point,
 theagent can examine different possible
sequences of actions, and choose the best
 This process of looking for the best sequence
is called search
 The best sequence is then a list of actions,

called solution
 Search algorithm
Defined as
 takinga problem
 and returns a solution

Once a solution is found

 theagent follows the solution
 and carries out the list of actions –
execution phase
Design of an agent
 “Formulate, search, execute”
Well-defined problems and solutions
A problem is defined by 5
components:
Initial state
Actions
Transition model or
(Successor functions)
Goal Test.
Path Cost.
Well-defined problems and solutions
A problem is defined by 4 components:
 The initial state
 that the agent starts in
 The set of possible actions
 Transition model: description of what each action
does.
(successor functions): refer to any state reachable
from given state by a single action
 Initial state, actions and Transition model define the
state space
 the set of all states reachable from the initial state by any
sequence of actions.
 A path in the state space:
 any sequence of states connected by a sequence of actions.
Well-defined problems and solutions
The goal test
 Applied to the current state to test
 if the agent is in its goal
-Sometimes there is an explicit set of possible goal
states. (example: in Amman).
-Sometimes the goal is described by the properties
 instead of stating explicitly the set of states
 Example: Chess
 the agent wins if it can capture the KING of the opponent on
next move ( checkmate).
 no matter what the opponent does
Well-defined problems and solutions
A path cost function,
 assigns a numeric cost to each path
 = performance measure
 denoted by g
 to distinguish the best path from others

Usually the path cost is

 thesum of the step costs of the individual
actions (in the action list)
Well-defined problems and solutions
Together a problem is defined by
 Initial state
 Actions
 Successor function
 Goal test
 Path cost function

The solution of a problem is then

 a path from the initial state to a state satisfying the goal
test
Optimal solution
 the solution with lowest path cost among all solutions
 Formulating problems
Besides the four components for problem
formulation
 anything else?
Abstraction
 the process to take out the irrelevant information
 leave the most essential parts to the description of the

states
( Remove detail from representation)
 Conclusion: Only the most important parts that are

contributing to searching are used

 Evaluation Criteria
formulation of a problem as search task
basic search strategies
important properties of search
strategies
selection of search strategies for
specific tasks
(The ordering of the nodes in FRINGE
defines the search strategy)
 Problem-Solving Agents
agents whose task is to solve a particular
problem (steps)
 goal formulation
 what is the goal state
 what are important characteristics of the goal state

 how does the agent know that it has reached the goal

 are there several possible goal states

 are they equal or are some more preferable
 problem formulation
 what are the possible states of the world relevant for solving
the problem
 what information is accessible to the agent

 how can the agent progress from state to state

Example
From our Example
1. Formulate Goal

- Be In Amman

2. Formulate Problem

- States : Cities
- actions : Drive Between Cities

3. Find Solution

- Sequence of Cities : ajlun – Jarash - Amman

 Our Example

1. Problem : To Go from Ajlun to Amman

2. Initial State : Ajlween

3. Operator : Go from One City To another .

4. State Space : {Jarash , Salat , irbed,……..}

5. Goal Test : are the agent in Amman.

6. Path Cost Function : Get The Cost From The Map.

7. Solution : { {Aj  Ja  Ir  Ma  Za  Am} , {Aj Ir  Ma  Za  Am} …. {Aj  Ja  Am} }

8. State Set Space : {Ajlun  Jarash  Amman}
 Example: Romania
On holiday in Romania; currently in Arad.
Flight leaves tomorrow from Bucharest

Formulate goal:
 be in Bucharest
Formulate problem:
 states: various cities
 actions: drive between cities

Find solution:
 sequence of cities, e.g., Arad, Sibiu, Fagaras, Bucharest



Example: Romania
 Single-state problem formulation
A problem is defined by four items:
1. actions or successor function S(x) = set of action–state pairs
 e.g., S(Arad) = {<Arad  Zerind, Zerind>, … }

2. goal test, can be

 explicit, e.g., x = "at Bucharest"
 implicit, e.g., Checkmate(x)

3. path cost (additive)

 e.g., sum of distances, number of actions executed, etc.
 c(x,a,y) is the step cost, assumed to be ≥ 0

A solution is a sequence of actions leading from the initial state to a

goal state

initial state e.g., "at Arad"

 Example problems
Toy problems
 those intended to illustrate or exercise
various problem-solving methods
 E.g., puzzle, chess, etc.

Real-world problems
 tend to be more difficult and whose
solutions people actually care about
 E.g., Design, planning, etc.
 Toy problems
Example: vacuum world
Number of states: 8
Initial state: Any
Number of actions: 4
 left, right, suck, noOp

Goal: clean up all dirt

 Goal states: {7, 8}

Path Cost:
 Each step costs 1
The 8-puzzle
 The 8-puzzle
States:
 a state description specifies the location of each of
the eight tiles and blank in one of the nine squares
Initial State:
 Any state in state space
Successor function:
 the blank moves Left, Right, Up, or Down
Goal test:
 current state matches the goal configuration
Path cost:
 each step costs 1, so the path cost is just the length
of the path
 The 8-queens
There are two ways to formulate the
problem
All of them have the common followings:
 Goal test: 8 queens on board, not attacking
to each other
 Path cost: zero
 The 8-queens
(1) Incremental formulation
 involves operators that augment the state
description starting from an empty state
 Each action adds a queen to the state

 States:
 any arrangement of 0 to 8 queens on board
 Successor function:
 add a queen to any empty square
 The 8-queens
(2) Complete-state formulation
 startswith all 8 queens on the board
 move the queens individually around

 States:
 any arrangement of 8 queens, one per column in
the leftmost columns
 Operators: move an attacked queen to a row,
not attacked by any other
 The 8-queens
Conclusion:
 the right formulation makes a big difference
to the size of the search space
 Example: River Crossing
Items: Man, Wolf, Corn, Chicken.
Man wants to cross river with all items.
 Wolf will eat Chicken
 Chicken will eat corn.

 Boat will take max of two.

Chapter -3 : Expert Systems
Basic Concepts of Expert Systems (ES)

Expertise
The extensive, task-specific knowledge that experts
possess.
The level of expertise determines the performance of
a decision.
Training, reading and experience in practice achieve
expertise.

34
Basic Concepts of Expert Systems (ES)
Features of ES
 Expertise: An EX must possess expertise that enables
it to make expert-level decision
 Symbolic reasoning: knowledge must be represented
symbolically
 Deep knowledge: concerns the level of expertise in a
knowledge base.
 Self-knowledge: must have a strong learning
capabilities to update their knowledge constantly (from
success, failure, and other knowledge resources).

35
Basic Concepts of Expert Systems (ES)
Why we need ES
1. Since experts may retire or leave, ES are an
excellent tool for preserving professional
knowledge crucial to a company's competitiveness
2. ES is an excellent tool for documenting
professional knowledge for examination or
improvement
3. ES is a good tool for training new employees and
disseminating knowledge in an organization
4. ES allow knowledge to be transferred more easily
at a lower cost

36
 Applications of ES
Classical successful ES
 DENDRAL: for molecular structure identification

 MYCIN: for medical diagnosis

 XCON: for configuration of the VAX computer

 Rule-based system

A system in which knowledge is represented

completely in terms of rules (e.g., a system
based on production rules)

37
 Applications of ES
Newer applications of ES
 Credit analysis systems : used in banks for risk

management assessment
 Pension fund advisors : help employees to plan their
retirement through their analysis that calculates their
pension benefits under different scenarios
 Automated help desks: enables small business to
deal with customer requests more efficiently
 Homeland security systems:
 Market surveillance systems

 Business process reengineering systems

38
 Applications of ES

Areas for ES applications

 Finance

 Data processing
 Marketing

 Human resources

 Manufacturing

 Homeland security

 Business process automation

 Health care management

39
 Structure of ES
ES environments:
1. Development environments
Parts of expert systems that are used by builders to
build the components and put knowledge into the
knowledge base.
It includes the knowledge base, the inference engine,
knowledge acquisition, and improving reasoning
capability.
The knowledge engineer and the expert are
considered part of these environments

40
 Structure of ES
2. Consultation environment
The part of an expert system that is used by a non-
expert to obtain expert knowledge and advice.
It includes the workplace (backboard), inference
engine, explanation facility, recommended action,
and user interface

41
Structure of ES

42
 Structure of ES

Three major components in ES are:

 Knowledge base

 Inference engine

 User interface

ES may also contain:

 Knowledge acquisition subsystem

 Blackboard (workplace)
 Explanation subsystem (justifier)
 Knowledge refining system

43
 Structure of ES

Knowledge acquisition (KA)

The extraction and formulation of knowledge derived
from various sources, especially from human experts
other sources: multimedia documents, textbooks,
DB, research reports, and the Web.
Knowledge base
A collection of facts, rules, and procedures organized
into schemas. The assembly of all the relevant
information and knowledge about a specific field of
interest

44
 Structure of ES
Inference engine
The part of an expert system (program) that provides a
methodology for reasoning about information in the KB
and on the blackboard for formulating conclusions.
User interfaces
The parts of computer systems that interact with users,
accepting commands from the computer keyboard and
displaying the results generated by other parts of the
systems

45
 Structure of ES
Blackboard (workplace)
An area of working memory set aside for the
description of a current problem and for recording
intermediate results in an expert system
Explanation subsystem (justifier)
The component of an expert system that can explain
the system’s reasoning and justify its conclusions

46
 Structure of ES
Knowledge-refining system
A system that has the ability to analyze its
own performance, learn, and improve itself
for future consultations

47
 How ES Work: Inference Mechanisms

Knowledge representation and organization

 Expert knowledge must be represented in a
computer-understandable format and organized
properly in the knowledge base
 Different ways of representing human knowledge

include:
 Production rules: easy to understand and adding
new rules is easy
 Semantic networks: A semantic network is a
knowledge structure that depicts how concepts
are related to one another and how
they interconnect
 Logic statements 48
 How ES Work: Inference Mechanisms
The inference process
Inference is the process of chaining
multiple rules together based on available
data to cover numerous conditions, since
expert knowledge can not be represented
in single rule.

49
 How ES Work: Inference Mechanisms

Popular approaches for inferencing:

 Forward chaining
A data-driven search in a rule-based system
 Backward chaining

A search technique (employing IF-THEN rules)

used in production systems that begins with
the action clause of a rule and works backward
through a chain of rules in an attempt to find a
verifiable set of condition clauses

50
 How ES Work: Inference Mechanisms

Development process of ES
 Is the process for eliciting knowledge from experts and
then storing the knowledge in the knowledge base.
A typical process for developing ES includes:
 knowledge acquisition
 Knowledge representation

 Selection of development tools

 System prototyping

 Evaluation

 Improvement

51
 Problem Areas Suitable for ES

Generic categories of ES

Interpretation Monitoring
Prediction Debugging
Diagnosis Repair
Design Instruction
Planning Control

52
 Development of ES
The development of ES includes:
1. Defining the nature and scope of the problem
 Rule-based ES are appropriate when the

nature of the problem is qualitative, knowledge

is explicit, and experts are available to solve
the problem effectively and provide their
knowledge

53
 Development of ES
The development of ES includes:
2. Identifying proper experts
A proper expert should have a thorough
understanding of:
 Problem-solving knowledge
 The role of ES and decision support technology

 Good communication skills

54
 Development of ES
The development of ES includes:
3. Acquiring knowledge
 Knowledge engineer
An AI specialist responsible for the technical
side of developing an expert system. The
knowledge engineer works closely with the
domain expert to capture the expert’s
knowledge in a knowledge base

55
 Development of ES
The development of ES includes:
4. Selecting the building tools
 There are three different kinds of
development tools:
1) General-purpose development environment
2) Expert system shell
A computer program that facilitates relatively
easy implementation of a specific expert
system. Analogous to a DSS generator

56
 Development of ES
There are three different kinds of development
tools:
3) Tailored turn-key solutions
 Contain specific features often required for
developing applications in a particular domain

57
 Development of ES
The development of ES includes:
5. Choosing an ES development tool
 Consider the cost benefits
 Consider the technical functionality and

flexibility of the tool

 Consider the tool's compatibility with the

existing information infrastructure

 Consider the reliability of and support from the

vendor

58
 Development of ES
The development of ES includes:
6. Coding the system
 The major concern at this stage is whether the
coding process is efficient and properly
managed to avoid errors
7. Evaluating the system
 Two kinds of evaluation:
 Verification:no error in the code and achieves
results the same as that acquired by the expert
 Validation: solve the problem correctly

59
Knowledge Acquisition (KA)
 Knowledge Acquisition
Steps in KA
• Data Gathering : determine what data will be collected, how,
and where this data is located
• Data Organising : rearrangement of the data gathered
• Summarising : raw data is converted to information form
which can be used for analysis
• Analysing : analysis of data to find new characteristics
• Synthesizing: joining statistics and patterns with a fixed
concepts
Benefits, Limitations, and Success Factors of
ES
Benefits of ES
 Increased output and productivity
 Decreased decision-making time

 Increased process and product quality

 Reduced downtime

 Capture of scarce expertise

 Flexibility

 Easier equipment operation

62
Benefits, Limitations, and Success Factors of
ES
Benefits of ES
 Elimination of the need for expensive
equipment
 Operation in hazardous environments

 Accessibility to knowledge and help desks

 Ability to work with incomplete or uncertain

information
 Provision of training

63
Benefits, Limitations, and Success Factors of
ES
Benefits of ES
 Enhancement of problem solving and decision
making
 Improved decision-making processes

 Improved decision quality

 Ability to solve complex problems

 Knowledge transfer to remote locations

 Enhancement of other information systems

64
Benefits, Limitations, and Success Factors of
ES
Problems with ES
 Knowledge is not always readily available
 It can be difficult to extract expertise from humans
 The approach of each expert to a situation assessment
may be different yet correct
 It is difficult to abstract good situational assessments
when under time pressure
 Users of ES have natural cognitive limits
 ES work well only within a narrow domain of
knowledge
 Most experts have no independent means of checking
whether their conclusions are reasonable
65
Benefits, Limitations, and Success Factors of
ES
Problems with ES
 The vocabulary that experts use to express facts and
relations is often limited and not understood by others
 ES construction can be costly because of the expense
of knowledge engineers
 Lack of trust on the part of end users may be a barrier
to ES use
 Knowledge transfer is subject to a host of perceptual
and judgmental biases
 ES may not be able to arrive at conclusions in some
cases
 ES sometimes produce incorrect recommendations

66
Benefits, Limitations, and Success Factors of
ES
Factors in disuse of ES
 Lack of system acceptance by users
 Inability to retain developers

 Problems in transitioning from development to

maintenance
 Shifts in organizational priorities

67
Benefits, Limitations, and Success Factors of
ES
ES success factors
 Level of managerial and user involvement
 Sufficiently high level of knowledge

 Expertise available from at least one

cooperative expert
 The problem to be solved must be mostly

qualitative
 The problem must be sufficiently narrow in

scope
68
Benefits, Limitations, and Success Factors of
ES
ES success factors
 The ES shell must be of high quality and
naturally store and manipulate the knowledge
 The user interface must be friendly for novice

users
 The problem must be important and difficult

enough to warrant development of an ES

 Knowledgeable system developers with good

people skills are needed

69
Benefits, Limitations, and Success Factors of
ES
ES success factors
 End-user attitudes and expectations must be
considered
 Management support must be cultivated
 End-user training programs are necessary
 The organizational environment should favor
adoption of new technology
 The application must be well defined,
structured, and it should be justified by
strategic impact

70
 ES on the Web
The relationship between ES and the
Internet and intranets can be divided into
two categories:
 The Web supports ES (and other AI)
applications
 The support ES (and other AI methods) give to

the Web

71