18

Architecture and Description Of Module Architecture and Description Of Module KNOWLEDGE BASE KNOWLEDGE BASE PRODUCTION RULES PRODUCTION RULES

Embed Size (px)

Citation preview

Page 1: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES
Page 2: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES
Page 3: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

The fundamental modules of an Expert System are:

Knowledge Base User Interface Inference Engine Explanation Facility Knowledge Acquisition Facility External Interface

Page 4: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

User

User Interface

Explanation Facility

Inference Engine

Knowledge Base

External InterfaceKnowledge

Acquisition Facility

Domain ExpertDBMS Sensors Communications

Architecture Of typical Expert System BACK

Page 5: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

• The core module of any expert system is its Knowledge-Base(KB). It is a warehouse of the domain-specific knowledge captured from the human expert via the knowledge acquisition module.

• There are many ways of representing the knowledge in the KB.

• Production rules are extremely popular KR structure today.•

BACK

Page 6: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

Sample rule 1.IF the marks obtained by the candidate is less than 50THEN declare him as failed

Sample rule 2.IF the car does not startTHEN there is fuel in the vehicleAND ignition system is working correctlyTHEN fault lies in fuel flow

Some production rules

A production rule states that a conclusion or an action or a consequent is sure to take place if the situation on the ‘IF’ part are satisfied, the rule is said to be triggered. If the action ‘THEN’ part of the rule is carried out, then the rule is said to have fired.

BACK

Page 7: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

Each rule provides a clear and convenient way of expressing modular chunks of knowledge.

Production rules with their modular and stylized coding mechanism helps in building an ES that has a higher degree of competence.

Integration of new rules into the knowledge base is easier because all one has to do is identify the location where new rule has to be placed.

Since each rule is a chunk of knowledge, they themselves are self-explanatory and explanation facilities can be easily built.

Stylized code of production rules also makes it feasible for manipulating individual rules with less impacts on other segments of the knowledge base.

A simple, straightforward inference mechanism can be easily implemented.

BACK

Page 8: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

Production rules lack expressive power for description of situations. While procedural knowledge can be implemented it is not that easy to make use of production rules for description.

When the number of rules is large, it becomes difficult to check whether or not a new rule brought into the system is redundant or in conflict with the existing ones. Consistency maintenance among rules is a tough task in ES.

Production rules have a fixed format. This requirement along with modularity and uniformity of the KB makes it necessary for one to state explicitly in the rule the entire context. When one starts specifying the context, it leads awkwardly long premises.

BACK

Page 9: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

User interface provides the needed facilities for the user to communicate with the system.

An user, normally would like to have a consultation with the system for the following aspects:

1.To get remedies for his problem.2.To know the private knowledge of the system.3.To get some explanations for specific queries.

  BACK

Page 10: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

Getting answers to specific queries forms the explanation mechanism of the expert system. Basically any user would like to ask the following basic questions “WHY” & “HOW”.

Conventional programs do not provide these facilities.

Explanation facility help the user in the following ways: 1.If the user is a domain expert, it helps in identifying what

knowledge is needed. 2.Enhances the user’s confidence in the system. 3.Serves as a tutor in sharing the system’s knowledge with the

user.

BACK

Page 11: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

Also called as ‘rule interpreter’ an inference engine, performs the task of matching antecedents from the user and firing rules.

Firing of rules causes two major things to happen. 1. Triggers another rule, thereby network of rule is triggered.

2. Implies that an action has been carried out. This adds new information to the database of inferred facts.

• Basically there are two approaches. They are: 1.FORWARD CHAINING 2.BACKWARD CHAINING

Page 12: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

• This works by matching the existing conditions of the problem with the antecedents of the rule in knowledge base.

• When the antecedent condition are met, the rule is fired and the new information derived is added to the concluded facts.

• Forward chaining is also known as data driven search or antecedent search.

• In fact, by triggering one rule, a series of rules are triggered. This method of matching only the if condition is called forward chaining.

FORWARD CHAINING

BACK

Page 13: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

• This is a reverse process of forward chaining.

• Hence the rule interpreter tries to match the ‘THEN’ condition instead of the ‘IF’ condition in the forward chaining.

• Because of this, backward chaining is also called goal driven search or consequent driven search.

BACK

Page 14: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

If one adopts a system with production rules and a ‘rule interpreter’ then the system is known as a production system.

One major issue to be sorted out in a production system is ‘conflict resolution’.

Conflict Resolution:-

Conflict resolution arises when there are more then one rule that can be fired in a situation and the

rule interpreter is to decide which rule to be fired, what is the order of triggering and whether to

apply all that are applicable or to be only selective.

Page 15: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

• There are a few strategies by which this can be resolved. Some of the conflict resolution strategies are :

1. Perform the first:- The system chooses the first rule that matches.

2. Sequencing technique:- Adopt the rules in the sequence they are.

3. Perform the most specific:- If there are two matching rules and one rule is more specific than the other, activate the more specific rule.

4. Most recent policy:- It is generally believed that a newly added rule is more knowledgeable than existing ones. Hence, if a system is adopting this method, it should fire the most recent ones.

BACK

Page 16: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

The major bottleneck in ES development is knowledge acquisition.

Present day ES do not have a sophisticated version of a learning system. Hence these systems perform by “being told”. To carry out the process of “being told”, systems provide what is called knowledge acquisition facility.

KAF creates a congenial atmosphere for the expert to share the expertise with the system.

BACK

Page 17: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

This provides the communication link between the ES and the external environment.

When there is a formal consultation, it is done via the user interface.

In real time expert system where they form the part of the closed loop system, it is not proper to expert human intervention every time to feed-in the conditions prevailing and get remedies. Moreover the time gap is too narrower in real time systems.

The external interface with its sensors gets the minute by minute information about the situation and acts accordingly.

Such real time ES will be of tremendous value in industrial process controls, in nuclear plants, in supersonic jet fighters etc.

BACK

Page 18: Architecture and Description Of Module Architecture and Description Of Module  KNOWLEDGE BASE KNOWLEDGE BASE  PRODUCTION RULES PRODUCTION RULES

BACK