新浪微博用户和信息的信用评估

Trust Evalution of Sina Weibo Users and Information

如何甄别社交网络中存在的不真实、误导性信息以及不可信的用户,建立微博中用户和信息的信用评价机制具有重要意义。目前流行的社交网络信用评价方法CoRank算法以国外的社交网络平台为研究对象。本文首先对比了新浪微博与国外主流社交网络的差别,在此基础上提出了针对新浪微博中用户和博文的信用评估方法---SWCoRank(Sina Weibo CoRank)算法。该方法引入“点赞”这一关系,定义了“带有点赞的复合动作”等复杂场景,讨论了这些场景在社交网络模型中所代表的语义以及相应的权重计算方法,利用上述权重计算微博用户和信息的信用,并针对其中的数据缺失问题提出相应的解决办法。最后,通过实际新浪微博数据验证了算法的有效性。

How to identify misleading information and untrustworthy users in social networks,and how to es-tablish a trust evaluation mechanism for users and information in microblogs is of great significance.At pre-sent,the popular social network trust evaluation method CORANK algorithm takes foreign social network platforms as the research object.This paper first compares the differences between Sina Weibo and foreign mainstream social networks,and on this basis,proposes a trust evaluation method for Sina Weibo users and blogs--SWCORANK(Sina Weibo CORANK)algorithm.The method introduced the relation"thumb up",de-fined the"composite action with the thumb up"and other complex scenario,discussed the semantics repre-sented by these scenarios in the social network model and the corresponding weight calculation method,used the above weight to calculate the trust of Weibo users and information,and put forward the corresponding so-lution to the problem of missing data.Finally,the effectiveness of the algorithm is verified by the actual Sina Weibo data.