Rules for implementing very large knowledge base systems

  • Authors:

  • J. B. Bocca;J. C. Freytag

  • Affiliations:

  • European Computer-Industry Research Centre, Arabellastr. 17, D-8000 Mfinchen 81, Federal Republic of Germany;European Computer-Industry Research Centre, Arabellastr. 17, D-8000 Mfinchen 81, Federal Republic of Germany

  • Venue:
  • ACM SIGMOD Record
  • Year:
  • 1989

Quantified Score

Hi-index 0.00

Visualization

Abstract

With the current research effort in knowledge base management systems (KBMSs) the interest in storing, retrieving, and processing rules efficiently, and using them for describing and implementing such systems has increased significantly. In this paper we contribute to this subject in two ways. First, we introduce Educe*, an integrated programming system for building large knowledge base systems. In particular, we describe the treatment of rules and complex logic terms, examine at the physical level the compilation of logic programs and its subsequent storage, retrieval and execution in the specified context, and discuss performance related aspects of our system.Second, we provide an overview on how to use rules for organizing the processing of user requests in a database/knowledge base environment. Specifically, we discuss aspects of rule-based query optimization and compilation. Finally, we outline additional query processing problems that could benefit from a rule-based specification and a rule-driven implementation.