一种参数可调的微博用户影响力评价方法

An Evaluation Method for Weibo User Influence with Adjustable Parameters

针对经典和改进的PageRank算法在计算微博转发和评论时采取PR均值分配进而造成PR值损失问题,提出了一种参数可调的微博用户影响力评价方法RCRank。RCRank在已有的PageRank算法基础上,增加了微博转发和评论影响微博用户影响力计算值的可调节参数。在实际数据集上的实验结果表明,在计算用户影响力时,相比于传统的PageRank和TunkRank算法,RCRank方法更接近真实情况。

Aiming at the problem of PR value loss caused by the classic and improved PageRank algorithm in the calculation of microblog forwarding and comments, a parameter-adjustable Weibo user influence evaluation method RCRank is proposed. Based on the existing PageRank algorithm, RCRank adds microblog forwarding and comments to adjust the parameters that affect the influence calculation of Weibo users. The experimental results on the actual data set show that the RCRank method is closer to the real situation than the traditional PageRank and TunkRank algorithms when calculating user influence.