Cogniitve Science : Expert System

Expert System :

1. Expert System is a set of program that manipulates encoded knowledge to solve problem in a specialized domain that normally requires human processing
2. An expert system is an intelligent program that solves problem in a narrow problem area by using high quality specific knowledge rather than algorithm
3. Expert system relays on the knowledge representation which is coded in suitable form for the system to use in its processes

Typical Task for Expert System :

1. Interpolation of data
2. Diagnosis of Malfunction(Medical)
3. Structural analysis ad configuration of complex objects
4. Predicting the future

Expert System Block Diagram : 

1. Knowledge Base : 
• The component that consist the system's knowledge
• A knowledge base consist of both declarative knowledge(facts about something) or procedural knowledge (information about something)
• Rule based production system is the currently used form of knowledge representation
• it stores complex structured and unstructured information used by computer
• it is mainly called as the database of information or knowledge
2. Inference Engine : 
• It directs the implementation of the knowledge and can be used to generate new information and solve problems
• They tends to apply new knowledge and deduced new knowledge
• It decides which heuristic search techniques are used to determine how the rues in the knowledge base are to be applied to the problem
• it directs expert system which rule are to be invoked, accessing the appropriate rules in the knowledge base, executing the rules and determining when an acceptable solution has been found
3. User Interface : 
• This component interfaces with user
• it consist of two basic components
• Interviewer Components : it mainly takes input of raw data from the user to be read into System
• Explanation Components : 
• It mainly gives the solution and also provides the steps of the system reasoning processes

Stages of Expert System : 

1. Identification : 
• Finding the characteristics of a problem
• Determining the exact nature of problem and state the precise goals that indicate exactly how we expect the expert system to contribute to the solution
2. Conceptualization : 
• Finding the primitive concepts and conceptual relations if the problem domain
• in this stage a problem is divided into the series of sub- problem and identified the relation among the pieces of each sub-problem
3. Formalization : 
• In this phase, the problem is connected to its proposed solution by analyzing the relationship deceptive in the conceptualization stage
• in this phase, a knowledge engineer must determine the tools that can greatly expedite the development processes, various techniques of knowledge representation and heuristic search
4. Implementation : 
• Actually formalized concepts are programmed onto the computer that is chosen for  development
• A prototype system is made which can be useful for evaluating progress but hardly a usable system
5. Testing : 
• it provides the opportunity to find the weakness in the structure
• it helps to implement a new extra features too

Features of Expert System : 

1. The program must be usable and useful
2. The program should be educational
3. The program should be able to explain its advice
4. The program should be able to respond the simple question
5. The program should be able to respond new knowledge
6. The program's knowledge should be easily modified.