融合博文内容和行为属性的Page Rank排序算法

发布时间：2018-06-01 01:23

  本文选题：微博 + 线性加权　； 参考：《科学技术与工程》2017年22期

【摘要】：针对当前微博影响力度量算法中多集中于用户行为属性,忽略博文、结点本身价值的问题,从微博用户信息出发,以线性加权模型为基础,综合分析用户的行为属性、博文相似度、节点相似度,创建影响力评价指标体系。利用Page Rank算法思想,提出了基于用户行为和博文内容的用户影响度量模型(user influence measurement rank,UMR)。通过采用新浪微博真实数据集测试,计算用户的影响力,验证了UMR算法在博文内容的基础上,能客观地反映用户的交互行为,消除僵尸用户对排序的影响,因而更科学、更合理。
[Abstract]:Aiming at the problem that the current micro - blog influence metric algorithm focuses on user behavior attributes , ignoring the value of the blog and the node itself , starting from the microblog user information , based on the linear weighted model , comprehensively analyzing the user ' s behavior attributes , the blog - text similarity , the node similarity , and creating the influence evaluation index system , and the page Rank algorithm is utilized to propose user influence measurement rank ( UMR ) based on user behavior and blog content . By adopting the real data set test of Sina Weibo , the influence of the user is calculated , and the UMR algorithm can objectively reflect the interaction behavior of the user on the basis of the content of the blog and eliminate the influence of the botnet user on the ordering , thus being more scientific and more reasonable .
【作者单位】： 江西理工大学信息工程学院;
【基金】：江西省研究生创新专项基金(YC2016-S316)资助
【分类号】：TP393.092
，

本文编号：1962290