引用本文
  • 卫玲蔚,胡斗,鲍祎楠,周薇,杨近朱,虎嵩林.基于时序特征和结构特征的社交网络谣言检测方法[J].信息安全学报,已采用    [点击复制]
  • WEI Lingwei,HU Dou,BAO Yinan,ZHOU Wei,YANG Jinzhu,HU Songlin.Jointly Exploiting Temporal and Structural Features for Rumor Detection on Social Media[J].Journal of Cyber Security,Accept   [点击复制]
【打印本页】 【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

过刊浏览    高级检索

本文已被:浏览 824次   下载 106  
基于时序特征和结构特征的社交网络谣言检测方法
卫玲蔚1,2, 胡斗3, 鲍祎楠1,2, 周薇1, 杨近朱1,2, 虎嵩林1,2
0
(1.中科院信息工程研究所;2.中国科学院大学网络空间安全学院;3.华北计算机系统工程研究所)
摘要:
随着社交网络的快速发展,越来越多的人在社交网络平台获取或分享信息。但是,在收获便利的同时,也为谣言提供了新的传播媒介。谣言的产生和传播严重降低了社交网络中的信息可信度,影响网络空间清朗环境的建设。现有基于深度学习的谣言检测模型往往基于内容特征或传播特征展开,而这些基于传播特征的模型要么只关注传播过程中的时序关系,或是仅挖掘谣言传播网络的结构特征来识别谣言,不能很好地学习一个全面的特征表示描述谣言传播过程中的时间和空间变化,限制了谣言检测的性能。针对此问题,本文提出一种通用的基于时序特征和结构特征的谣言检测方法,共同探索谣言传播过程中的时间模式与传播树结构特征,学习一个全面的谣言特征表示,提高谣言检测的性能。为了评估模型的有效性,利用3个公开的真实数据集,对本文提出的模型进行分析验证。实验结果表明,在谣言传播的各个阶段,该方法均能够有效地检测出社交网络中的谣言。
关键词:  谣言检测  时序特征  传播结构特征  门控循环单元  图卷积神经网络  社交网络
DOI:10.19363/J.cnki.cn10-1380/tn.2023.08.05
投稿时间:2020-12-25修订日期:2021-03-04
基金项目:国家重点研发计划(No.2018YFC0806900)
Jointly Exploiting Temporal and Structural Features for Rumor Detection on Social Media
WEI Lingwei1,2, HU Dou3, BAO Yinan1,2, ZHOU Wei1, YANG Jinzhu1,2, HU Songlin1,2
(1.Institute of Information Engineering,Chinese Academy of Sciences;2.School of Cyber Security, University of Chinese Academy of Sciences;3.National Computer System Engineering Research Institute of China)
Abstract:
With the rapids development of the social network, more and more people obtain or share information on social network platforms. Unfortunately, the convenient environment of social network platforms also provides rumors a new propaga-tion medium. The existing deep learning-based rumor detection models have been developed based on content character-istics or propagation characteristics including temporal features and structural features. However, most of these models either only model the temporal information in rumor propagation or only focus on the network structure features of rumor propagation to identify rumors, which cannot learn a comprehensive eigenvector representation well and limit the per-formance of rumor detection. Aiming at this problem, we propose a novel rumor detection model, to jointly model both structural features and temporal patterns in the rumor propagation. Accordingly, the model can learn a comprehensive representation of rumor characteristics. In addition, the model can effectively alleviate the time mode distortion caused by pruning. Extensive experiments on three real-world datasets demonstrate that the proposed model can improve the performance of rumor detection.
Key words:  rumor detection  temporal information  propagation structural information  gated recurrent unit  graph convolutional network  social network