Trust-Enhanced Recommender Systems for Efficient On-Line Collaboration

摘要: Trust has been explored by many researchers in the past as a solution for assisting process of recommendation production. In this work we are examining feasibility building networks trusted users using existing evidence that would be provided standard recommender system. As there is lack models today could help finding relationship between trust and similarity build our own uses set empirical equations to map metrics into Subjective Logic trust. paper perform evaluation proposed model being part complete Finally, present interesting results from shows performance benefits modeling technique well its impact on user community it evolves over time.