Twitter的Follow关系和Retweet关系对比

Comparison between Follow and Retweet on Twitter

研究Twitter在线社交网络中,Follow关系和Retweet关系在传播用户影响力和表征用户同质性这两方面的差异。为研究两者在传播用户影响力上的差异,定义了V f变量和V r变量分别度量Follow关系和Retweet关系的作用;为研究两者在表征用户同质性上的差异,分别基于Follow关系和Retweet关系构造出对应的社交网络图,并采用wvRN算法分别对两个网络内的用户进行分类。通过对比用户的V f变量值和V r变量值发现,Retweet关系在传播用户影响力方面的作用优于Follow关系;通过对比分类结果发现,Follow关系比Retweet关系更能表征用户的同质性,基于Follow关系的分类精度比基于Retweet关系的分类精度高20%,分类结果同时揭示不同类别的用户体现出了不同的关注和信息互动特性。基于上述研究说明Follow关系和Retweet关系所携带的信息是不同的。

This work aimed to compare the differences between Follow relationship and Retweet relationship in propagating the Twitter users influence and measuring the Twitter users homophily. In order to find their differences in the first aspect, this paper defined two verities： Vf was defined for the Follow relationships and Vr was for the Retweet relationships. In order to analyze the differences in measuring the homophily, it built two social network separately based on Follow relationships and Retweet relationships. Then it adopted the wvRN algorithm to classify users in these two networks. By comparing the Vf and Vr values for a group of user, it can be found that the Retweet relationship win in propagating the users influence. By comparing the classification results, it can be found that the Follow relationship contains more homophily information, that the accuracy based on Retweet is 20% higher than that based on Retweet. But the results also illustrates that different kinds of users have different communication characteristics. This research shows that Follow relationships and Retweet relationships contain different information.