基于异质信息网络表征学习的微博虚假信息甄别研究  

Research on Weibo Misinformation Detection Based on HeterogeneousInformation Network Representation Learning

作　　者：王世雄[1] 吴泽政 Wang Shixiong;Wu Zezheng(School of Economics&Management,Zhejiang Sci-Tech University,Hangzhou 310018)

机构地区：[1]浙江理工大学经济管理学院,杭州310018

出　　处：《情报杂志》2024年第12期152-160,共9页Journal of Intelligence

基　　金：国家社会科学基金一般项目“智能生成信息迷雾问题的形成机理与治理机制研究”(编号:24BGL114)研究成果。

摘　　要：[研究目的]社会网络的飞速发展与突发公共卫生事件的时有发生,使得大量的、具有迷惑性的虚假信息混杂社交媒体中,甄别此类信息已成为用户信息素养的重要组成部分。[研究方法]以微博上的异质信息网络为研究对象,综合考虑微博文本的语义特征和用户行为的非语义特征,引入多头注意力机制融合生成集成表示实现虚假信息甄别,并从信息内容、参与用户、用户与信息交互三个维度进行特征挖掘与量化分析。[研究结论]研究表明,基于异质信息网络表征学习的虚假信息甄别方法具有较好的实用性,有助于解构虚假信息的特征,为突发公共卫生健康事件的虚假信息治理及辩症施策提供有益参考。[Research purpose]With the rapid development of social networks and the occurrence of public health emergencies,a large amount of deceptive misinformation is often intertwined within social media platforms.Identifying such misinformation has become an important component of users'information literacy.[Research method]This study focuses on heterogeneous information networks on Weibo.By considering both the semantic features of Weibo texts and the non-semantic features of user behaviors,a multi-head attention mechanism is introduced to integrate generated ensemble representations for misinformation detection.Feature mining and quantitative analysis are conducted from three dimensions:information content,participating users,and user-information interactions.[Research conclusion]Research indicates that the misinformation detection method based on heterogeneous information network representation learning demonstrates practicality.It contributes to deconstructing the characteristics of misinformation,thereby providing beneficial insights for governing and strategizing against misinformation during public health emergencies.

关 键 词：异质信息网络 表征学习 突发公共卫生事件 虚假信息 用户行为 社交媒体 微博文本 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]