Abstract: Based on the traditional PageRank algorithm, a new algorithm model is present to assess the influence of bloggers in community of microblog. The influence of bloggers in the microblog community reflects its authority, is one of the core concept of microblog community research. Through statistic analysis about the network characteristics on average degree, clustering coefficient and average path length indicators, which verify microblog community networks have a “Small World” significant features. Through two indicators of activity of users and blog quality, bloggers impact assessment system is constructed. Introduced the factor of the spread ability of the bloggers, the idea of PageRank algorithm is used to design a new influence ranking algorithm model （Influence Rank） to assess the blogger influence. Through the experimental comparison, it finds that Influence Rank algorithm in addition to considering the relationship of the node also takes into account the characteristics of the node itself, which can more accurately and objectively reflect the blogger’s influence ranking.

Key words: community of microblog, influence of bloggers, PageRank, Influence Rank

摘要： 提出了基于传统的PageRank算法的改进模型评估微博社区博主的影响力。微博社区中博主的影响力反映其话语权的大小，是研究微博社区的核心概念之一。通过对平均度、聚类系数和平均路径长度等网络特征指标的统计，验证了微博社区网络具有“小世界”的显著特性。从用户活跃度和博文质量两个角度出发，构建了博主影响力的评价指标，引入了博主传播能力这个因子，利用PageRank算法的思想设计了新的影响力排名（Influence Rank）算法模型来评估博主影响力。通过实验对比发现Influence Rank算法在考虑节点间的关系之外还考虑了节点本身的特性，能够更加准确客观地反映博主的影响力排名。

关键词: 微博社区, 博主影响力, PageRank算法, Influence Rank算法