Conversational Expert System

Abstract

Expert Systems, a computer program that uses artificial intelligence techniques to solve problems within a specialized field that usually requires human experience. Expert systems rely on two components: the knowledge base and the inference engine. Knowledge Base is an organized set of facts about the scope of the system. The inference engine interprets and evaluates the facts in the knowledge base in order to provide an answer. Typical functions of expert systems include classification, diagnostics, monitoring, design, scheduling and planning of specialized endeavors. A knowledge base for a system contains thousands of rules. Probability is often attached to each output or output from the system, because the conclusion is not certain. For example, a system for the diagnosis of eye diseases, based on the information provided to it, may indicate a 90% probability that a person has a blueness and may also include conclusions with lower probability. An expert system may also show the sequence of rules through which it has come to an end; tracking this flow helps the evaluator evaluate the credibility of conclusions and recommendations and also serves as a learning tool

for students.

Human experts often use heuristic rules, or "thumb rules", as well as simple production rules, such as rules derived from geometric books. Thus, the credit manager of a particular insurance company may know that an applicant with a bad credit record, but who has a clean

record since acquiring a new job, may actually have a good credit risk.