1 / 27

Chapter 5

Artificial Intelligence 605451. Chapter 5. Expert System. Dr.Hassan Al-Tarawneh. Human Expert Problem Solving. An expert is someone who possesses specialized knowledge about a particular domain This type of knowledge is called domain knowledge

oldakerm
Télécharger la présentation

Chapter 5

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Artificial Intelligence 605451 Chapter 5 Expert System Dr.Hassan Al-Tarawneh

  2. Human Expert Problem Solving • An expert is someone who possesses specialized knowledge about a particular domain • This type of knowledge is called domain knowledge • The expert’s domain knowledge is stored in his Long Term Memory (LTM)

  3. How does an Expert Provide An Advise to Solve a Problem? • The expert obtains facts about the problem (case facts) from the advisee. • The case facts are stored in the expert’s shot term memory (STM) • The expert then reasons about the problem by combining the STM facts with the LTM knowledge • Using this process, the expert infers new problem information and eventually arrives at a conclusion regarding the problem

  4. How does an Expert Provide An Advise to Solve a Problem?

  5. An Expert Solving a Problem • Assume you have a problem with your car. You take it to an auto mechanic (an expert) • The mechanic stored in his LTM knowledge needed for diagnosing a variety of problems with cars • You tell the mechanic the problem with the car, i.e., “the car would not start” • The mechanic stores this information 9cae facts) in his STM and begins to reason with it • Using the case facts along with his domain knowledge the mechanic might infer that “the problem is in the electrical system” • The mechanic then adds this belief to his STM and continues to reason about the problem • The mechanic then conducts battery tests to support his belief and concludes that “the fault is a bad battery” • Using this approach the mechanic cannot only solve the problem but also explain the reasoning behind his conclusions

  6. Expert System • Expert systems solve problems using a process that is similar to the methods used by human experts

  7. Architecture of an Expert System

  8. Modules of an Expert System – The user Interface • The interaction between an expert systems and a user if facilitated by the User Interface • The interaction between the expert system and the user should be • Highly interactive • In a natural language style • Simple and comfortable • The basic requirement of an user interface is to obtain reliable information from the user by asking questions • User interfaces are designed by employing a variety of interface styles • Menus • Question and answer screens • Natural language • Graphics

  9. Modules of an Expert System – The Knowledge Base • Knowledge base is the heart of the expert system which contains the domain knowledge. i.e. the problem solving knowledge about a particular domain • The typical way of representing knowledge is an expert system is in the form of IF…THEN… rules • The IF… THEN … rule logically relates information contained in the IF part to other information contained in the THEN part

  10. Modules of an Expert System – The Knowledge Base • Example : Two rules to represent the knowledge for automobile fault diagnosis : • RULE 1 • IF the car will not start • THEN the problem may be electrical system • RULE 2 • IF the problem may be electrical system • AND the battery voltage is below 10 volts • THEN the fault is a bad battery

  11. Modules of an Expert System – The Knowledge Base • Since, an expert system at a time addresses problem related to one domain only, therefore, the knowledge base at a time contains problem-solving knowledge about a single, particular domain • Structurally, knowledge base’s are generic in nature and can be used with other expert system’s inference engine

  12. Modules of an Expert System – The Inference Engine • The inference engine is the processor is an expert system that applies the knowledge in the knowledge base to derive solutions • The inference engine models the process of human reasoning to solve a problem • The inference engine matches the facts contained in the working memory with the domain knowledge contained in the knowledge base, to draw conclusions about the problem.

  13. How Does the Inference Engine Work? • The inference engine searches the rules for a match between the rule premises and the information contained in the working memory • When a match is found, the inference engine adds the rule’s conclusion to the working memory. • The inference engine continues to scan the rules looking for new matches

  14. How Does an Expert System Works? Knowledge Base • RULE 1 • IF the car will not start • THEN the problem may be electrical system • RULE 2 • IF the problem may be electrical system • AND the battery voltage is below 10 volts • THEN the fault is a bad battery

  15. Step 1 • Expert System : • Does the car not start? • USER : • YES • The user asserts this fact info the working memory. Since this new fact supports RULE 1, the expert system asserts into the working memory the conclusion of RULE 1. • USER ASSERTS • The car will not start • SYSTEM ASSERTS • The problem may be in the electrical system

  16. Step 2 • Expert System : • Is the battery voltage below 10 volts? • USER : • YES • The working memory now contains the facts that support RULE 2 and the expert system asserts its conclusion into the working memory • USER ASSERTS • The battery voltage is below 10 volts • SYSTEM ASSERTS • The fault is a faulty battery

  17. Modules of an Expert System – The Knowledge Base Editor • The knowledge base editors are used to edit existing knowledge base • Knowledge base editors can assist : • The addition of new knowledge • The maintenance of rules and their syntax • To perform consistency checks on the updated knowledge base

  18. Building an Expert System : The Knowledge Engineering Process • The main interest of expert system design lie in the Problem’s Knowledge • Expert system designers, therefore, acquire, organize, and study the knowledge to gain an understanding of the problem • The overall process of developing an expert system is termed as knowledge engineering

  19. People Involved in the Development of an Expert System • Domain Expert • Knowledge Engineer • End - User

  20. Role of the Domain Expert • The domain expert is a person who : • Possesses the skills and knowledge to solve problems in a specific domain • Understand the domain’s problem solving techniques, such as shortcuts, handling imprecise information, evaluating partial solutions and other skills • Has a lot of experience in the domain that is targeted by the expert system • Recognizes important pieces of information and uses them to his/her advantage to arrive efficiently at a solution

  21. Role of the Domain Expert • The domain expert is a person who : • Provides the knowledge of the problem area to the expert system developer • Has good communication skills so that he/she can pass the knowledge to the knowledge engineer • Has the time to devote towards the development of the expert system

  22. Role of the Knowledge Engineer • The knowledge engineer is a person who designs, builds and tests the expert system • The knowledge engineer is first responsible for knowledge acquisition from the domain expert.

  23. Role of the Knowledge Engineer • Knowledge acquisition is achieved through • Reading introductory texts, journals, or manuals about the problem domain • Interviews with the domain expert: the knowledge engineer tried to extract the deeply embedded problem solving methods to the domain expert • Giving the domain expert a few test cases and asking him to explain the solution process

  24. Role of the Knowledge Engineer • Human experts are unreliable when it comes to explaining their problem solving methods – they may skip some details which are basic to them or forget to mention some steps that they believe are too obvious • Such information may seem trivial to the domain expert but for a novice and expert system developer this is valuable information • It is therefore the job of the knowledge engineer to ask for simple, basic details and steps, and try to extract information as much information as possible from the domain expert. • Knowledge engineer must therefore also have good communication skills

  25. Role of the Knowledge Engineer • After knowledge acquisition the job of the knowledge engineer is to design the expert system in AI terms • Typical responsibilities include : • Select a knowledge representation scheme – rules, frames, etc. • Determine the inference strategy – forward or backward chaining • Design the most friendly user interface possible • Choose the development platform • Choose the development software

  26. Role of the Knowledge Engineer • The knowledge engineer is responsible for • Coding • Testing • Revising the system until it is acceptable to the users • Integrating the expert system into the intended workplace • Maintaining the system after completion

  27. Role of the End User • The end user is the individual who would eventually use the expert system • End users determine the major design constraints, such as • The level of explanation • How information is entered – question & answer, menus, list, etc. • How should the expert system be accessed • The form of the final results • The various utility programs needed – text editors, spell checkers, calculators • The user interface • The end user is very important in the initial stages of the development life cycle as it provides a broad description of the problem

More Related