一种基于社交网络舆论影响的推荐算法

摘要

社交网络中的朋友间信任度是影响用户决策的重要因素之一.基于社交网络的拓扑结构和相邻节点间的信任度,构建推荐模型,给出推荐算法,并进行理论分析和仿真模拟.研究结果显示,社交网络结构和信任度在社会交往的作用下会对用户的产品预期产生较大的影响.社交网络中单个产品的推荐者数量随时间呈S型曲线变化.推荐系统中产品推荐等级序列呈尾部截断的幂律分布.该研究不但分析了用户对产品预期的形成机理,而且推荐算法在一定程度上提高了推荐的效率.%The trust among friends is an important factor that affects users making decision in social network.Based on the topology of social network and trust degree of neighbor node, the recommendation system is constructed in this paper, and the recommendation algorithm is also presented.Then, some theoretical analysis and simulations have also been made out.The results show that the topology and trust degree in social network will have a great impact on user's anticipation for product during the process of social communication.The mount of recommender for single product in social network varies with time by following S-curve and the rank of recommended products obeys the power-law distribution with cut-off tail.This research not only analyzes the forming mechanism of user's anticipate, but also to some extent enhances the efficiency of recommendation by using the recommendation algorithm.