Fuzzy user preference model for top-k search

The task of modeling complex user preferences face the problem of understandability to common users and the problem of querying the dataset for preferred objects. We propose natural and complex model of user preferences decomposed with respect to particular attribute values. Partial preferences are combined by monotone aggregation function. The user model is stored in ontology structure. We also present an extension of top-k objects search algorithm to provide a query evaluation of the proposed model.