社交网络虚假舆情检测研究进展

Research on False Public Opinion Detection in Social Network

虚假舆情的传播极大地影响了网络空间的有序健康发展,而社交网络的广泛使用在很大程度上加剧了虚假舆情的恶意注入与传播.虚假舆情是一个社会科学概念,在自然科学的研究领域缺少具体的外延.具体化了虚假舆情的表现形式,将误导信息、假新闻、谣言、讽刺、欺凌、恶意言论等拓展为虚假舆情的外延表现,以便从自然科学角度分析虚假舆情的注入与传播.同时,为了更清晰有效地刻画早期检测的概念,将舆情传播周期的概念融入虚假舆情检测,分析了在各个阶段进行虚假舆情检测研究的区别和侧重点.系统地总结了近年来在虚假舆情检测方面的突出研究,主要介绍了特征工程、技术路线、可用数据集以及评价指标等,并讨论了虚假舆情检测技术面临的研究与挑战.

The spread of false public opinion has hardly affected the orderly and healthy development of cyberspace,and the widespread use of social networks has exacerbated the malicious injection and spread of false public opinion.False public opinion is a social science concept that lacks specific extension in the research field of natural science.This paper specified the manifestation of false public opinion and expanded misleading information,false news,rumors,sarcasm,bullying,and malicious language into the extension of false public opinion to facilitate the analysis of the injection and dissemination of false public opinion from the perspective of natural science.The concept of the public opinion transmission cycle was integrated into false public opinion detection to more clearly and effectively depict early detection.The differences and emphasis of false public opinion detection research at various stages were analyzed.This paper summarized the outstanding research on false public opinion detection in recent years using a systematic literature review method,introduced feature engineering,technical route,available data sets,and evaluation indicators,and discussed the research and challenges faced by false public opinion detection technology.