摘要
[目的/意义]探索社交媒体环境下辟谣信息发布有效性的影响因素及不同影响因素之间的交互作用机理。[方法/过程]从辟谣主体、辟谣信息和辟谣情景3个维度构建辟谣信息发布有效性评估指标体系,提取认知值、转发影响力等5个指标对辟谣有效性进行综合评价;通过Python爬取微博辟谣数据,以指标体系中影响因素为特征,辟谣有效性为标签,建立8种机器学习分类模型进行实证分析,并利用SHAP可解释性模型探讨最佳LGBM分类模型的影响因素及其交互作用机制。[结果/结论]附带真相图片等因素对辟谣信息发布有效性有显著正面影响;纯文本类型等因素对辟谣信息发布有效性有显著负面影响;不同辟谣主题下不同因素对辟谣信息发布有效性的影响机制存在显著差异。
[Purpose/significance]This study explores the influencing factors of the effectiveness of rumor-refuting information release in the social media environment and the interaction mechanism among different influencing factors.[Method/process]This paper constructs an evaluation index system for the effectiveness of rumor-refuting information release from three dimensions:the rumor-refuting subject,the rumor-refuting information,and the rumor-refuting scenario,extracts five indicators such as cognitive value and forwarding influence to conduct a comprehensive evaluation of the effectiveness of rumor-refuting,crawls the data of refuting rumors on Weibo through Python,establishes eight machine learning classification models for empirical analysis by taking the influencing factors in the index system as characteristics and the effectiveness of refuting rumors as labels,and uses the SHAP interpretability model to explore the influencing factors and their interaction mechanisms of the best LGBM classification model.[Result/conclusion]Factors such as attaching truth pictures have a significant positive impact on the effectiveness of rumor-refuting information release;factors such as the type of plain text have a significant negative impact on the effectiveness of rumor-refuting information release;there are significant differences in the influence mechanisms of different factors on the effectiveness of rumor-refuting information release under different rumor-refuting themes.
作者
陈建文
周珊
赵文军
Chen Jianwen;Zhou Shan;Zhao Wenjun(School of Mathematics and Statistics,Hunan University of Technology and Business,Changsha Hunan 410205;School of Management Science and Engineering,Hunan University of Technology and Business,Changsha Hunan 410205)
出处
《情报探索》
2025年第6期33-40,共8页
Information Research
基金
湖南省教育厅科学研究项目“突发网络公共事件中情感传播机理及引导策略研究”(项目编号:22A0427)
湖南省哲学社会科学基金一般项目“突发事件下社会心理健康预警监测研究”(项目编号:23YBA167)的成果之一。
关键词
社交媒体
辟谣微博
信息发布
有效性评估
SHAP可解释性
social media
rumor-refuting weibo
information release
effectiveness evaluation
SHAP interpretability
作者简介
陈建文(1977-),男,博士,教授,硕士生导师,研究方向为文本挖掘与数据挖掘;周珊(1998-),女,2022级硕士研究生,研究方向为文本挖掘与数据挖掘;赵文军(1981-),男,博士,教授,硕士生导师,研究方向为网络信息行为。
