随着近年来互联网技术的快速发展, 应用获取平台都面临着信息过载的问题. 面对大量应用, 解决用户不能快速准确地找到满足其偏好的应用的问题迫在眉睫. 已有的如Cosine、Pearson等协同过滤方法普遍存在稀疏性、冷启动和可扩展性等问题, 从而对推荐结果产生影响. 文章在考虑用户社 交关系、偏好及信任关系的基础上, 提出了融合用户社交关系与信任关系的应用推荐方法. 该方法融合用户社交关系、点赞与标签等特征及其对应用的偏好计算相似度, 并基于好友的信任关系与用户声誉计算信任度, 最终将相似关系与信任关系进行有效结合, 实现应用推荐. 实验结果表明: 所提应用推荐方法能更好的体现用户间的社交与信任关系, 且能有效提高应用推荐的有效性与准确度.

With the rapid development of Internet technology in recent years, application acquisition platforms are facing the problem of information overload. Facing a large number of applications, it is urgent to solve the problem that users cannot quickly and accurately find applications that meet their preferences. Existing collaborative filtering methods such as Cosine and Pearson generally have problems such as sparsity, cold start up, and scalability, which affect the recommendation results. In this paper, based on the user's social relationship, preference and trust relationship, an application recommendation method that integrates the user's social relationship and trust relationship is proposed. This method combines the characteristics of users' social relationships, likes and tags and their preferences for applications to calculate the similarity, and calculates the degree of trust based on the friend's trust relationship and the user's reputation, and finally effectively combines the similarity relationship and trust relationship to achieve application recommendation. Experimental results show that the application recommendation method proposed in this paper can better reflect the social and trust relationship between users, and can effectively improve the effectiveness and accuracy of application recommendation.