USING A RELATIONAL DATABASE TO SUPPORT EXPLANATION IN A KNOWLEDGE-BASED SYSTEM

Explanation is important in knowledge-based systems (KBS's) for two reasons: it informs the user of the method used in producing the solutions to a problem thereby increasing his/her confidence in the system, and it helps the knowledge engineer refine the problem solving knowledge. In KBS applications where the number of facts in the knowledge base is very large, the amount of information that is needed for effective explanation support can become too voluminous to be stored in main memory. In this paper, we present an approach to model and manage the information needed for explanation using a relational database. We show how different types of explanation can then be produced by appropriate queries on this database. We formulate representative queries for some major types of explanation using ESQL, an extension of SQL.