A prioritized "and" aggregation operator for multidimensional relevance assessment

Abstract

In this paper a new model is proposed for aggregating multiple criteria evaluations for relevance assessment based on a refinement of the "min" ("and") operator. The peculiarity of such an operator which also distinguishes it from the traditional "min" aggregation operator is that the extent to which the least satisfied criterion plays a role in determining the overall satisfaction degree depends both on its satisfaction degree and on its importance for the user. If it is not important at all, its satisfaction degree is not considered, while if it is the most important criterion for the user, only its satisfaction degree is considered (like with the traditional "min" operator). The usefulness and effectiveness of such a model are demonstrated by means of a case study on personalized Information Retrieval with multicriteria relevance. Some preliminary experimental results are also reported. © Springer-Verlag 2009.