Disjunctive Deductive Databases: Semantics, Updates, and Architecture

Proceedings of the Fourth Bar-Ilan Symposium on Foundations of Artificial Intelligence

Volume

Proceedings of the Fourth Bar-Ilan Symposium on Foundations of Artificial Intelligence

Contents

The basic assumption in relational and deductive databases is that there are no gaps in our knowledge. That is, the database cannot store data that contain null values or data that is indefinite. In practical situations knowledge is not precise, and there are gaps in our knowledge. These gaps may be due to null values in the data, may arise when we combine several databases that lead to inconsistent theories, or may occur because information is indefinite in nature, such as in military or medical applications. In this paper we describe semantics for disjunctive deductive databases that extend the work in deductive databases, solve the view update problem, and permit indefinite data to be represented efficiently. Hierarchic, stratified, and normal stable models of disjunctive databases are described. An architecture is proposed for a disjunctive deductive database system and a class of theories for which the architecture will be effective is discussed.