Expert Systems
- The expert Systems are computer programmes created to address challenging issues in a certain field at a degree of competence and intellect beyond that of a typical human.
- Expert System Features
- Excellent work
- Understandable
- Reliable
- Extremely accommodating
- Expert Systems’ Capabilities
- Advising
- guiding and supporting people as they make decisions
- Demonstrating
- Finding a resolution
- Diagnosing\sExplaining
- Analyzing input
- Forecasting outcomes
- Rationalizing the outcome
- Offering substitute solutions for an issue
- They are not able to
- replacement of human decision-makers
- Having human abilities
- generating precise results given a limited information base
- enhancing their own understanding
- Expert System Components
- Knowledge Base
- Interface of the Inference Engine User
- I The Knowledge Base
- It includes excellent, domain-specific expertise. To demonstrate intellect, one must be knowledgeable. The acquisition of extremely precise and accurate knowledge is crucial to the success of any ES.
- Knowledge
- Facts are gathered to create the data. The data and facts regarding the task domain are how the information is arranged. When facts, information, and prior experience are gathered together, the result is knowledge.
- Knowledge Base Components
- An ES’s knowledge base is a repository for both factual and heuristic information.
- Factual Knowledge: This is the data that academics and knowledge engineers in the task domain generally agree upon.
- Heuristic Knowledge: It involves experience, precise judgement, assessment skills, and guesswork.
- Knowledge illustration
- It is the process by which the knowledge in the knowledge base gets arranged and formalised. It takes the form of rules that are IF-THEN-ELSE.
- Acquiring Knowledge
- The quality, accuracy, and completeness of the data kept in the knowledge base play a key role in the success of any expert system.
- The Knowledge Engineers, along with a variety of professionals and scholars, contribute readings to the knowledge base. A knowledge engineer is someone who possesses empathy, rapid learning, and case analysis abilities.
- By interviewing, recording, and seeing the topic expert at work, among other methods, he gathers information from him. The input is then meaningfully categorised and arranged by him into IF-THEN-ELSE rules that the interference machine may employ. The knowledge engineer keeps an eye on the ES’s advancement as well.
- (ii) Engine of Inference
- In order to derive an accurate, perfect solution, the Inference Engine must employ effective processes and principles.
- When using knowledge-based ES, the Inference Engine pulls information from the knowledge base and applies its own logic to arrive at a specific solution.
- Regarding rule-based ES
- repeatedly applies rules to the facts—which are derived from previous application of the rules—to the facts.
- if necessary, adds additional information to the knowledge base.
- resolves conflicts between rules when more than one rule applies to a specific situation.
- The Inference Engine employs the following techniques to suggest a fix:
- Chaining Forward
- Reverse Chaining
- Chaining Forward
- An expert system uses this tactic to respond to queries about “What can happen next?”
- In this case, the Inference Engine deduces the result by following the series of conditions and derivations. Before coming to a conclusion, it filters through all of the rules and facts.
- When working on a conclusion, consequence, or effect, this method is used. For instance, forecasting the state of the share market in response to interest rate fluctuations.
- Reverse Chaining
- An expert system uses this tactic to determine the reason behind the incident.
- The Inference Engine looks for circumstances that might have existed in the past to produce this outcome based on what has previously occurred. This method is used to determine the reason or cause. For instance, the identification of human blood cancer.
- User Interface (iii)
- User interface facilitates communication between the ES’s user and the device. Generally, it is Natural Language Processing so that users with extensive knowledge of the task topic can use it. It’s not required for the ES user to be an artificial intelligence specialist.
- It provides an explanation of how the ES reached at a specific recommendation. The following formats for the explanation are possible:
- On-screen natural language displayed.
- oral narratives in everyday speech.
- Rule numbers are listed and shown on the screen.
- It is simple to determine whether the deductions are credible thanks to the user interface.
- Conditions for an Effective ES User Interface
- It need to assist users in achieving their objectives as quickly as feasible.
- It ought to be made to accommodate users’ preferred or current work habits.
- The technology need to adjust to the needs of the user, not the other way around.
- It ought to utilise user input effectively.
- Limitations of Expert Systems
- No technology can provide a comprehensive and simple answer. Large systems are expensive, time-consuming to design, and resource-intensive in terms of computers. There are certain limitations associated with ESs.
- Technology’s limitations
- acquiring knowledge that is difficult
- ES are challenging to keep up.
- high expenses for development
- Expert System Applications
- Where ES can be used is given in the table below.
- Application Description: Design domain: automotive and camera lenses.
- Medical Domain Diagnosis Systems to perform human medical procedures and infer the source of sickness from observable data.
- Systems of Monitoring
- data is constantly compared to the system as it is being observed or to the recommended behaviour, such as leak detection in a lengthy petroleum pipeline.
- Systems for Process Control
- using monitoring to drive physical process control.
- Domain of Knowledge
- identifying hardware issues with computers and cars.
- Finance/Commerce
- Expert System Technology: potential fraud detection, suspicious transaction analysis, stock market trading, airline and freight scheduling
- There are various ES technology levels available. Among the technologies for expert systems are
- I Environment for Expert System Development:
- Tools and hardware are components of the ES development environment. They are as follows:
- Mainframes, minicomputers, and workstations.
- Advanced Symbolic Computing Programming languages like LISP (LISt Programming) and LOGique (PROgrammation) (PROLOG).
- big databases.
- (iii) Instruments:
- They significantly lower the time and expense needed to create an expert system.
- strong editors and multi-window debugging tools.
- They offer quick prototyping.
- possess built-in definitions for the inference design, knowledge representation, and models.
- (iii) Shells:
- Without a knowledge base, an expert system is all that a shell is. Developers can access an explanation facility, user interface, inference engine, and knowledge acquisition through a shell. A few shells, for instance, are listed below:
- The Java Expert System Shell (JESS) offers a fully functional Java API for the development of expert systems.
- A shell called Vidwan was created in 1993 at Mumbai’s National Centre for Software Technology. It makes it possible to encode knowledge using IF-THEN rules.
- Creating Expert Systems: Basic Procedures
- Iterations are used in the creation of ES. The following are steps in creating the ES:
- Step 1: Determine the Issue Domain
- An expert system can only address problems that are appropriate for them.
- Locate the subject matter experts for the ES project.
- Determine whether the system is cost-effective.
- Step 2: Create the System
- Determine the ES Technology
- Determine the level of integration with the databases and other systems.
- Recognize how the concepts can most effectively convey the domain knowledge.
- Step 3: Utilizing the knowledge base, develop the prototype The expert’s domain knowledge is acquired by the knowledge engineer through work.
- Make use of If-THEN-ELSE rules to express it.
- Step 4: Evaluate and Enhance the Model
- Sample cases are used by the knowledge engineer to test the prototype and look for any performance issues.
- The ES prototypes are tested by end users.
- Step 5: Create and finish the ES test and make sure the ES interacts with databases, end users, and other information systems, among other components of its environment.
- Thoroughly record the ES project.
- Educate the user on ES.
- Step 6: Keep the ES in Place
- Update and evaluate the knowledge base frequently to keep it current.
- As other information systems change, provide support for new interfaces.
- Expert Systems’ Advantages
- Availability: Because software is produced in large quantities, it is readily available.
- Reduced Production Cost: The cost of production is fair. They are therefore reasonably priced.
- Speed: They provide excellent speed. They lessen the effort that a person exerts.
- Reduced Error Rate: When compared to human mistake, the error rate is minimal.
- Lowering Risk: They are able to operate in hazardous environments for people.
- Consistent response: They operate without becoming agitated, tense, or exhausted.