基于社交网络和地理位置的用户关系预测

The Prediction of User Relationship Based on Social Network and Geographic Location

为了解决社交网络用户关系预测的不精确问题,通过采用时空分析方法对移动通信用户的轨迹和通话关系数据进行分析,研究了结合地理和时间的重合度以及基于用户时空特征权重赋值算法,并提出了基于社交网络和地理位置相似度的用户关系预测方案。经过实验证明,提出的算法能够改善社交网络预测的精度。

In order to solve the problem of inaccurate prediction of the user relationship in social networks, the trajectory of mobile user and communication relation data were analyzed based on time-space analysis method. The weight assignment algorithm combined with the geographical and temporal consistency and user＇s temporal-spatial characteristics was investigated. A prediction scheme on user relationship based on the similarity of the social network and geographical location was proposed. Experiments demonstrate that the proposed algorithm can improve the prediction accuracy of social networks.