软件导刊2025，Vol.24Issue(6)：24-31,8.DOI:10.11907/rjdk.241657
在线创新社区用户交互信息深度与广度识别
Recognition of Information Depth and Breadth of User Interaction in Online Innovation Communities:A Method Using LDA Topic Model
摘要
Abstract
Describing user interaction characteristics through external attributes of online interaction is not comprehensive enough.Therefore,a method using text mining technology to identify the structural features of interaction information based on user interaction content is pro-posed.Firstly,based on the knowledge structure theory,a research framework is constructed to identify the themes and their probability distri-butions involved in interactive texts using LDA topic models;Secondly,based on the topic distribution of user interaction documents,K-means clustering is performed on the documents to divide the different knowledge domains covered by user interaction;Finally,by construct-ing a user document the second mock examination network,we can describe the scope and depth of user interaction in online innovation com-munity,and then identify the depth and breadth of information.Research on Microsoft Power BI community shows that the model can identify 45 topics and 18 knowledge domains involved in user interaction.It was found that 78%of users only participate in shallow interactions in a single domain,and the depth and breadth need to be improved.It can also identify generalist,specialized,and cross domain specialized us-ers,providing guidance for online innovation communities to analyze user interaction characteristics and identify key users in various fields.关键词
在线创新社区/用户交互/文本挖掘/信息深度/信息广度/微软Power BI社区/LDA主题模型Key words
online innovation community/user interaction/text mining/information depth/information width/Microsoft Power BI commu-nity/LDA topic model分类
信息技术与安全科学引用本文复制引用
张欣茹,曹鑫磊,陈佳丽,侯延香..在线创新社区用户交互信息深度与广度识别[J].软件导刊,2025,24(6):24-31,8.基金项目
山东省自然科学基金青年项目(ZR2021QG014) （ZR2021QG014）
山东省人文社会科学项目(2023-JCXK-007) （2023-JCXK-007）
