Ranking Semantic Web Services Using Rules Evaluation and Constraint Programming

Abstract

Current Semantic Web Services discovery and ranking proposals are based on user preferences descriptions whose expressiveness are limited by the underlying logical formalism used. Thus, highly expressive preference descriptions, such as utility functions, cannot be handled by the kind of reasoners traditionally used to perform Semantic Web Services tasks. in this work, we outline a hybrid approach to allow the introduction of utility functions in user preferences descriptions, where both rules evaluation and constraint programming are used to perform the ranking process. Our proposal extends the Web Service Modeling Ontology with these descriptions, providing a highly expressive framework to specify preferences, and enabling a more general ranking process, which can be performed by different engines.