关键词: 社交网络, 双向推理算法（TWIT）, 信任值推断, 机器学习

Abstract: Online social network is one of the most successful services in mobile Internet. How much one can trust others has been one of users’ most concerned questions. To avoid private information being exposed by malicious users, a two-way algorithm called TWIT is proposed, to calculate local trust values. It does the inference from the view of both trustable and untrustable users. Taking the transferability of the friends relationships into account, MTWIT algorithm is proposed based on the idea of machine learning, to eliminate the restriction of the precondition that there must exists a path between any two users, and take the advantage of inferring trust values with higher accuracy in the meantime. Moreover, it can adapt to dynamic topology in social network. The experimental results on datasets from Sina microblog fully illustrate this conclusion.

Key words: social network, Two-Way algorithm to Infer Trust（TWIT）, trust inferrence, machine learning