We compare two approaches to the maintenance of derivable data: a relational approach based on materialized views and a semantic approach based on declarative assertions. In particular we examine whether they can support completeness of data sets. None of these approaches is able to support adequate automatic actions if a database system perceives data incompleteness, because these systems do not have knowledge about the involved real world. Therefore a solution should rely on the ability of users to specify adequate actions. An essential part of the solution is to specify the required cardinality of a data set and to compare it with the actually calculated cardinality. In the case of a conflict, the involved user must be informed about the cause of the conflict and this user must execute some adequate action(s) until the conflict is resolved.

Keywords: Assertions, Data Completeness,

Intelligent Databases, Materialized Views.

Publication Data

Proceedings of the Sixth IASTED International Conference on Artificial Intelligence and Soft Computing (ASC 2002), Banff, Canada, pp. 113-117.

Editor: H. Leung

Publisher: ACTA Press, Anaheim, Calgary, Zürich

Year: 2002

ISBN: 0-88986-346-6