基于加权转发关系的社交网络意见领袖识别算法  

Opinion Leaders Based on Forwarding Relationship Identification and Analysis of Social Network

作　　者：仇丽青[1] 范鑫 QIU Li-qing;FAN Xin(School of Information Science and Engineering,Shandong University of Science and Technology,Qingdao 266590,China)

机构地区：[1]山东科技大学信息科学与工程学院,山东青岛266590

出　　处：《软件导刊》2018年第7期111-115,共5页Software Guide

基　　金：国家自然科学基金项目(61502281)

摘　　要：在社交网络中进行意见领袖的识别对信息传播分析、舆情监测、网络营销等有着重要意义。目前,很多挖掘意见领袖的研究仅基于简单的粉丝关注或转发方式,而忽视了用户关系中的权重因素,不符合信息传播的实际情况。因此,提出加权网络下基于微博转发关系的FW-Rank(Forwarding Weight Rank)算法。该算法使用新浪微博数据进行实验,相比于通过简单关注关系形成社交网络的方式,FW-Rank算法的最终结果更加合理准确。Identification of opinion leaders in social networks has great significance on the analysis of information dissemination,public opinion monitoring,network marketing and so on.At present,most of the research on opinion leaders in social network is merely focused on fans attention or forwarding method,and ignores the weight factors in the user relationship,which makes the final result unreasonable.Therefore,the paper proposes a novel algorithm called FW-Rank which takes the forwarding relationship of the network into account.At last we conduct the experiments on Sina MicroBlog dataset.Compared to the baseline algorithms,the final results of the proposed algorithm is more reasonable and accurate.

关 键 词：社交网络 意见领袖 微博 FW-Rank 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]