A new approach to the filtering of ill-known data

In this paper, the issue of querying databases containing ill-known values is addressed. A new type of query is introduced based on criteria applying to the representations of ill-known data. In the regular possibilistic framework, the only authorized selection criteria concern the value that an item can take. We introduce a new querying framework in order to exploit some concepts which are part of the model used for representing ill-known data, and which concern the qualification of imprecision/uncertainty. The representation-based querying framework defined in this paper constitutes the first step to the introduction of an explicit manipulation of the concepts of imprecision and/or uncertainty into a database query language