Query processing in temporal evidential databases

Abstract

This paper first presents the temporal evidential database (TED) model, which incorporates time into the evidential relational database model. Evidential databases support both uncertain and imprecise information using Dempster-Shafer's (D-S) theory of evidence (or evidence. theory). The TED model also adopts an evidence theoretic approach to managing both relative and absolute times under uncertainty. The query language for the TED model is then presented. It is shown how to integrate the evaluations of both temporal and non-temporal predicates in query processing. This not only presents a solution in the TED model, but also points out this query processing problem is common in the database models that aim at managing both time and uncertainty. The query language also has a novel modal syntax with respect to evidential functions.