Semiring-based trust evaluation for information fusion in social network services

As online social network services (SNS) are booming and gaining tremendous popularity, there is a sharply increasing amount of information exchange and interactions among SNS users. Taking this advantage, users in SNS make decisions via collecting and combining information from different sources (i.e. other users). However, there exists a large variance of trustworthiness among SNS populations, which is threatening the quality of the information fusion process. In such circumstances, trust relationships among users in SNS are very important in decision making as well as for the success of many SNS-based applications, e.g. recommender systems and ad targeting. An appropriate trust inference mechanism for trust evaluation is necessary in extending the knowledge base of trust opinions and tackle the issue of limited trust information due to link sparsity in social networks. In this work, we model the trust relationship among users in SNS as a 2-dimensional vector, and propose a semiring-based model for trust propagation and fusion as the building block of our trust inference framework. Specifically, in our approach, both trust and distrust (i.e., positive and negative trust) are both considered, and opinion conflict resolution is supported by our framework of trust inference.