LBSN中结合相遇和拓扑结构的朋友推荐算法

New meeting-sensitive and network-sensitive algorithm for friend recommendation in LBSN

利用基于位置的社交网络(LBSN)中的共享位置信息,提出一种结合相遇和拓扑结构的朋友推荐算法.该方法定义用户关系为相遇模型,通过考虑朋友间时间和空间的相遇特征,基于随机路点模型计算轨迹间在相同时间下的相遇频率作为用户相似度,结合拓扑相似度作为最终的推荐依据.实验结果表明,所提出的方法较传统基于拓扑的好友推荐算法准确率更高.

Based on the shared location information in Location-Based Social Network(LBSN),a friend recommendation algorithm based on geographical neighbors is proposed.The method first defines the user relationship as the encounter model,considering the encounter characteristics of time and space between friends,and then calculates the encounter frequency between the trajectories at the same time based on the random waypoint model as the user similarity,and finally combines the topological similarity as the final recommended basis.The experimental results on the real data set show that the proposed method is more accurate than the traditional topology-based friend recommendation algorithm.