基于PageRank算法的双维度微博用户影响力实时度量模型  

Real-time Measurement Model of the Influence of Micro-blog Users with Double Dimensions Based on Page Rank Algorithm

作　　者：李志宏[1] 庄云蓓 

机构地区：[1]华南理工大学工商管理学院,广东广州510640

出　　处：《系统工程》2016年第2期128-137,共10页Systems Engineering

基　　金：国家自然科学基金重大资助项目(71090403)

摘　　要：基于PageRank算法的微博用户影响力度量模型存在着"与粉丝数量相关度过高"和"未考虑用户的个人传播意愿"两大不足。为此通过综合用户个人的传播能力和传播意愿两大维度,同时在模型中引入时间变量,提出了微博用户影响力动态度量(UI)模型。研究结论表明:(1)基于UI模型的影响力值不再与粉丝数量相关;(2)微博网络用户行为具有同好性,且传播能力的时间间隔服从幂律分布;(3)UI模型具有较好的实时性及收敛性。To overcome the two short comings of measurement models of users＇ influence-- ＂high correlation with the number of fans＂ and ＂lacking considering the users＇ spreading desire＂, based on PageRank algorithm, this paper improves PageRank algorithm based on two dimensions of ＂users＇ spreading ability＂ and ＂users＇ spreading desire＂ by introducing time variable into the model, a new model for measuring users＇ influence--UI model is built. The experiment result shows that： （1） the new model overcomes the above measured insufficiency; （2） the interval of users＇ behavior follows power-law distribution and has ＂homoplasy and convergence＂ property; （3） the UI model is of great real-time and great convergence.

关 键 词：影响力度量 PAGERANK算法 UI模型 传播能力 传播意愿 实时性 

分 类 号：G206[文化科学—传播学]