Integrity Constraint and Rule Maintenance in Temporal Deductive Knowledge Bases

摘要

The enforcement of semantic integrity constraints in data and knowledge bases constitutes a major performance bottleneck. Integrity constraint simplification methods aim at reducing the complexity of formula evaluation at run-time. This paper proposes such a simplification method for large and semanti-cally rich knowledge bases. Structural, temporal and asser-tional knowledge in the form of deductive rules and integrity constraints, is represented in Telos, a hybrid language for knowledge representation. A compilation method performs a number of syntactic, semantic and temporal transformations to integrity constraints and deductive rules, and organizes simplified forms in a dependence graph that allows for efficient computation of implicit updates. Precomputation of potential implicit updates at compile time is possible by computing the dependence graph transitive closure. To account for dynamic changes to the dependence graph by updates of constraints and rules, we propose efficient algorithms for the incremental maintenance of the computed transitive closure.