Abstract: Relying solely on the similarity to locate neighbor users has a serious negative impact on the performance of traditional collaborative filtering algorithm. Importing the trust mechanism in social networks, a model is built up from the perspective of individual’s subjective trust and global reputation in social circle. Direct trust is formed by considering users’ interactions, rating differences and preferences. Indirect trust can be generated by reputation and expert prior trust model. Then the proposed algorithm constructs the trust relationship among users through a dynamic weighting manner. The parameter η is introduced to coordinate double attributes including trust and similarity to make the user relationship more closer, which can solve the problems of new users as well as sparsity effectively. Pragmatic analysis reveals that the improved model has the remarkable results.

Key words: subjective trust, global reputation, user preferences, expert prior trust

摘要： 仅凭相似度来定位邻居用户对传统协同过滤算法的性能有严重的负面影响。引入社会网络中的信任机制，从个体在社交圈中的主观信任和全局声誉角度出发建模。分别考虑用户交互、评分差和用户偏好调节生成直接信任度。利用声誉及专家信任优先模型聚合生成间接信任度，将两者动态加权形成用户之间的信任关系。用参数[η]协调信任和相似双属性，使用户关系更加紧密，有效地解决新用户和稀疏性问题。经实证，改良后的模型颇有成效。

关键词: 主观信任, 全局声誉, 用户偏好, 专家优先信任