Aggregates in possibilistic databases

Abstract

Fuzzy set theory represents a uniform framework for extending the relational database model to han-, dle imprecision of information found in the real world. None of the existing proposals for data models handling imprecision has dealt with queries involvin aggregate operators. This paper presents a framewor for ek handling aggregates in the context of imprecise information. Two kinds of aggregates, namely, scalar aggregates and aggregate functions, are being supported. We consider three cases: aggregates within approximate queries on precise data, aggregates within precisely specified queries on possibilistic data, and aggregates within vague queries on imprecise data. These extensions are based on fuzzy set-theoretical concepts such as the extension principle and the possibilistic expected value.