Modelling uncertainty in multimedia database systems: an extended possibilistic approach

  • Authors:

  • Guy De Tré;Rita De Caluwe

  • Affiliations:

  • Ghent University, Dept. of Telecommunications and Information Processing, Computer Science Laboratory, Sint-Pietersnieuwstraat, 41 B-9000 Gent, Belgium;Ghent University, Dept. of Telecommunications and Information Processing, Computer Science Laboratory, Sint-Pietersnieuwstraat, 41 B-9000 Gent, Belgium

  • Venue:
  • International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

In database systems one often has to deal with constraints in order to compel the semantics of the stored data or to express some querying criteria. This is especially the case for multimedia database systems where information of different media types, as e.g. sound, video, images and texts, needs to be managed. A problem with the modelling of (the uncertainty about) the degree of satisfaction of a constraint, is the handling of missing information. In this paper, it is shown how extended possibilistic truth values can be employed to cope with this problem. The notion of an extended possibilistic truth value has been obtained from the assumption that the truth value, which expresses the degree of satisfaction of a constraint, can be undefined. This is for example the case if the constraint cannot be evaluated due to the non-applicability of (some of) its elements. An illustrative database definition and database querying example is presented and discussed.