共同好友推荐java_一种社交网络下的好友推荐算法

A New Link Prediction Algorithm Based on Social Network

ZHANG Mengzhe

1

张萌哲 女 硕士研究生，主要研究方向：社交网络和个性化推荐

ZHANG Chengwen

2

张成文，男，副教授，主要研究方向：云计算和个性化推荐.

1、School of computer science, Beijing University of Posts and Telecommunications, Beijing 100876

2、School of computer Beijing Key Laborary of Intelligent Telecommunications Software and Multimedia, Departmentof Computer, Beijing University of Posts and Telecommunications.

Abstract：The traditional social network friends recommendation algorithm based on relation information evaluates user similarity by the number of the common followers and followees between the two users, ignoring the closeness between the individuals of the followees and the user. To solve this problem , this paper focus on the user's interaction information to compute the intimacy of the individuals and the user, and a new link prediction algorithm based on PageRank is proposed, which integrates the interaction information and relation information. The experiments are performed on the Tencent microblog data, and the results show that the algorithm combines of the two kinds of information can achieve the higher recall rate and precision rate.