A Collaborative-Filtering Recommendation Based on the Trust Network

Abstract

Many recommendation algorithms are proposed for the internet business system. However, many researches are focus on the model of users, items and their relation-ships. Recently, some studies proposed to use the trust relationships among users to optimize the accuracy of recommendation. But many studies recently did not consider the indirect trust relation between users. In our manuscript, we proposed an hybrid approach to mine the indirect and direct trust relation between any two users in the data set. Based on experiments used the real world public data set, our algorithm outperforms the other state-of-the-art approaches recently in the accuracy.