情报杂志2025，Vol.44Issue(5)：122-129,8.DOI:10.3969/j.issn.1002-1965.2025.05.015
主题模型网络分析:计算传播视野下的舆情文本挖掘新进路
Analysis of Topic Model Networks:A New Approach to Complex Text Mining from the Perspective of Computational Communication
摘要
Abstract
[Research purpose]This paper aims to propose a new complex text mining method,"The Analysis of Topic Model Net-works"(ANTMN),to solve the limitations of traditional public opinion text mining methods in terms of too strong motivation,too much manual intervention,rough themes,etc.,and provide effective method support for public opinion governance,risk identification and pub-lic opinion guidance.[Research method]The limitations of traditional methods and the academic progress of computational transforma-tion were analyzed through literature review,and the innovative advantages of ANTMN compared with existing methods were demonstrated through comparative analysis.[Research result/conclusion]The research finds that ANTMN follows the three-step method of"subject modeling","network analysis"and"community detection",obtains objectivity with"pure induction orientation",and surpasses the stat-ic limitation of traditional methods with"mixed computing orientation".ANTMN also shows strong practical application value in the re-search of public opinion communication,that is,it puts forward public opinion communication problems in empirical phenomena,analyzes public opinion communication mechanism in data-driven analysis,and constructs public opinion communication system in trend predic-tion.关键词
舆情传播/主题模型网络分析/舆情文本挖掘/计算传播Key words
public opinion dissemination/analysis of topic model networks/public opinion text mining/computational communication引用本文复制引用
张媛,刘兴澳..主题模型网络分析:计算传播视野下的舆情文本挖掘新进路[J].情报杂志,2025,44(5):122-129,8.基金项目
贵州社会科学重大课题"贵州提升思想引领力、舆论影响力、文明塑造力、文化竞争力的问题及对策研究"(编号:24GZZB15) （编号:24GZZB15）
西安交通大学基本科研业务费(人文社科类)自由探索学生类项目"青少年焦虑情绪的智能识别与疏导研究"(编号:SK2025041)研究成果. （人文社科类）
