一种新的微博社区用户影响力评估算法  

A NEW ASSESSMENT ALGORITHM ON INFLUENCE OF USERS IN MICROBLOG COMMUNITY

作　　者：刘玲[1] 杨长春[1] 

机构地区：[1]常州大学信息科学与工程学院,江苏常州213164

出　　处：《计算机应用与软件》2017年第7期212-216,261,共6页Computer Applications and Software

基　　金：国家自然科学基金项目(61272367);江苏省产学研前瞻性联合研究项目(BY2014037-08)

摘　　要：近年来,微博用户都凭借其自身在社区中的影响力来对信息传播做出贡献,尤其是活跃的大V用户能够引起信息广泛的传播。为了在微博社区中提高用户影响力衡量的准确性,提出了一种基于传统的PageRank算法和用户交互行为的用户影响力改进算法(IUIR算法),此算法通过直接质量指数和间接质量指数来构建微博用户的质量指数,再结合近期用户的活跃度来构造用户影响力评价公式。在新浪微博数据集上进行实验,并与传统的PageRank算法作比较,结果表明,该算法能够更有效地反映微博用户影响力的排名。In recent years, Microblog users made contributions to the dissemination of information by virtue of its own influence in the mieroblog community, especially active users could spread information widely. In order to improve the accuracy of measuring the influence of users in mieroblog community, an improved influence algorithm based on traditional PageRank algorithm and user interactions （IUIR algorithm ） is proposed. This algorithm created the final quality index of users through direct and indirect mass index, combining with the recent user activity to construct user influence assessment formula. Through the experimental analysis on the Sina mieroblog data collection to compare this proposed algorithm with the traditional PageRank algorithm, the results indicate that the algorithm could reflect the users＇ influence ranking more accurately.

关 键 词：微博社区 用户影响力 用户质量指数 近期活跃度 PAGE RANK 算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]