DATABASE CONCEPTS FOR THE SUPPORT OF KNOWLEDGE-BASED SYSTEMS

The purpose of a database management system is the reliable and efficient management of shared persistent data. Current knowledge-based systems are designed for representing and processing knowledge in a way that is cognitively adequate. Since knowledge should be kept persistent as well, it seems to be interesting to combine the techniques of database management systems and knowledge-based systems into something like a knowledge base management system. We shortly show some conceptual differences between these two types of systems and then discuss how database concepts should be evolved in order to meet the requirements of knowledge-based systems. Using LILOG-DB as an example, we demonstrate in detail the modeling capabilities a database system should provide in order to support typical knowledge structures. Especially, we discuss the design of FLL, a first-order logic based query language supporting an order-sorted complex-object data-model under an Open-World Assumption.