PageRank模型的改进及微博用户影响力挖掘算法

IMPROVEMENT OF PAGERANK MODEL AND MINING ALGORITHM OF MICROBLOG USER INFLUENCE

随着Web技术的发展,微博逐渐成为当下最流行的社交平台之一。微博中用户影响力计算是相关研究中的焦点问题。通过对PageRank模型的改进,提出一种新的用户影响力挖掘算法PR4WB(PageRank for Micro Blogs),解决了传统的PageRank算法由于页面权威值的等分传递带来的潜在误差过大的问题。PR4WB算法在考虑微博中用户关系的同时,利用社会网络概念将自身的活跃度、博文质量及可信性加以关联,形成动态的评价模型。基于Twitter数据的实验表明,PR4WB算法能更加准确、客观地反映出用户的实际影响力。

With the development of Web technology , microblog has become one of the most popular social platforms.The calculation of user influence in microblog is the focus of related research. Through the improvement of the PageRankmodel , a new user influences mining algorithm PR4WB （PageRank for Microblog） is proposed to solve the problem that the traditional PageRank algorithm has too much potential error due to the transfer of page authority value. PR4WBalgorithm takes into account the user relationship in microblog while using the concept of social network to link itsactivity , blog quality and credibility to form a dynamic evaluation model. Experiments based on Twitter data show that ,PR4WB algorithm can more accurately and objectively reflect the user, s actual influence.