开源社区中的意见领袖识别及跟随效应仿真  

Opinion Leader Identification and Following Effect Simulation in the Open Source Community

作　　者：徐冰村 张晓冬[1] 

机构地区：[1]北京科技大学经济管理学院

出　　处：《情报理论与实践》2019年第12期101-107,共7页Information Studies:Theory & Application

基　　金：国家自然科学基金项目“智能制造环境下生产系统的组织集成建模与协同仿真研究”的成果之一,项目编号:71871018

摘　　要：[目的/意义]有效识别开源社区中的意见领袖,模拟开源社区中用户跟随行为,提高合作效率.[方法/过程]根据开源社区的评论关系构建知识协作网络,选取社会网络分析中的4个重要评价指标进行分析.再采用基于TOPSIS意见领袖综合评价方法对网络中关键意见领袖节点进行识别,应用改进的SIR模型模拟开源社区中用户跟随行为.[结果/结论]通过分别与由结构洞确定和灰色关联法确定的意见领袖在网络中引起的传播情况比较,识别出的意见领袖在传播率和传播时间等各方面都更能体现意见领袖的特点.[局限]SIR跟随效应模型没有体现社交网络中具体点对点的传播过程.[Purpose/significance]To identify opinion leaders effectively,simulate user follow-up behaviour and improve collaboration efficiency in the open source community.[Method/process]According to the comment relationship of the open source community,the knowledge collaboration network is constructed,and four important evaluation indicators in the social network analysis are selected for analysis.The key opinion leader nodes in the network are identified using the TOPSIS-based comprehensive evaluation method of opinion leaders.Applying an improved SIR model to simulate user follow-up behavior in the open source community.[Result/conclusion]By comparing the communication generated by opinion leaders determined by structural holes and gray relative analysis method respectively,the opinion leaders identified in this paper can better reflect the characteristics of opinion leaders in terms of transmission rate and time of communication.[Limitations]The SIR following effect model can,t reflect the specific point-to-point propagation process in social networks.

关 键 词：开源社区 意见领袖识别 SIR模型 跟随效应仿真 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]