Towards Context-Enriched Trust Prediction: A Proposal

This short paper describes an early stage of research aiming at improving trust prediction in social networks. It builds on recent findings on observed correlation between trust and user similarity. A machine-learning approach to the trust prediction problem with use of contextual information is proposed and experimental work envisaged. An overview of the existing results is presented and it indicates that the proposed approach constitutes a novel combination of ideas and, as such, has a potential to contribute to the previous research in the area.