融合评论序列二义性与生成用户隐私特征的谣言检测

doi:10.11772/j.issn.1001-9081.2023081176

《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (8): 2342-2350.DOI: 10.11772/j.issn.1001-9081.2023081176

融合评论序列二义性与生成用户隐私特征的谣言检测

孟文凡¹, 周丽华¹(), 王晓旭²

^1.云南大学信息学院，昆明 650504
^2.云南大学滇池学院，昆明 650228

收稿日期:2023-08-31 修回日期:2023-09-14 接受日期:2023-10-09 发布日期:2024-08-22 出版日期:2024-08-10
通讯作者: 周丽华
作者简介:孟文凡（1996—），男，贵州独山人，硕士研究生，CCF会员，主要研究方向：数据挖掘、信息扩散、谣言检测
周丽华（1968—），女，云南华坪人，教授，博士，CCF会员，主要研究方向：数据挖掘、多视角学习、社会网络分析 lhzhou@ynu.edu.cn
王晓旭（1995—），女，安徽阜阳人，讲师，硕士，主要研究方向：数据挖掘、大数据。
基金资助:
国家自然科学基金资助项目(62062066);云南省基础研究计划项目(202201AS070015)

Rumor detection by fusing ambiguity in comment sequences and generating user privacy features

Wenfan MENG¹, Lihua ZHOU¹(), Xiaoxu WANG²

^1.School of Information Science and Engineering，Yunnan University，Kunming Yunnan 650504，China
^2.Dianchi College of Yunnan University，Kunming Yunnan 650228，China

Received:2023-08-31 Revised:2023-09-14 Accepted:2023-10-09 Online:2024-08-22 Published:2024-08-10
Contact: Lihua ZHOU
About author:MENG Wenfan ，born in 1996，M. S. candidate. His researchinterests include data mining， information diffusion， rumor detection.
ZHOU Lihua ，born in 1968，Ph. D.， professor. Her researchinterests include data mining， multi-view learning， social networkanalysis.
WANG Xiaoxu，born in 1995，M. S.， lecturer. Her researchinterests include data mining， big data.
Supported by:
This work is partially supported by National Natural ScienceFoundation of China （62062066）； Yunnan Basic Research Program（202201AS070015）.

摘要/Abstract

摘要：

现有谣言检测工作存在以下问题：1）没有同时捕获评论序列的文本语义特征和时间周期特征；2）在隐私保护环境下无法获取用户个人资料，导致传播结构中的信息难以充分融合。为此，提出融合评论序列二义性与生成用户隐私特征的谣言检测模型（RD-CSGU）。综合考虑了评论序列不同视角下的文本语义特征和时间周期特征，同时构建了反映传播过程中用户之间社交互动关系的谣言传播异质网络，并基于该网络中的语义关系通过生成对抗网络（GAN）生成用户的隐私特征，解决了用户个人资料访问受限的问题。在Twitter15、Twitter16、Weibo数据集上展开有效性验证，与次优基线模型GLAN（Global-Local Attention Network）相比，RD-CSGU的准确率（Acc）分别提升了0.9、2.2和1.8个百分点，真谣言F1（TR-F1）值分别提升了2.6、6.8和1.9个百分点；结合消融实验及GAN生成嵌入分析的实验结果表明，RD-CSGU能有效检测出社交媒体平台上发布的谣言帖子。

关键词: 谣言检测, 评论序列, 传播异质网络, 生成特征, 传播结构

Abstract:

There are some problems in existing rumor detection works， such as not fully integrating the information within propagation structure because of the deficiency of simultaneously capturing text semantic features and time periodic features in comment sequences and the inability to access the user personal profiles in a privacy-protected environment. To address the above problems， a Rumor Detection model fusing ambiguity in Comment Sequences and Generating User privacy features （RD-CSGU） was proposed. Text semantic features and time periodic features from different perspectives of comment sequences were comprehensively considered. Meanwhile， a heterogeneous network of rumor propagation for describing the social interaction relationship among users during the propagation process was constructed， based on which user privacy features were generated through a Generative Adversarial Network （GAN） based on the semantic relationships， overcoming the limitation of user personal profiles. The effectiveness of the proposed model was validated on Twitter15， Twitter16 and Weibo datasets. Compared with the suboptimal baseline model GLAN （Global-Local Attention Network）， RD-CSGU achieved improvements of 0.9， 2.2 and 1.8 percentage points in Accuracy （Acc）， as well as improvements of 2.6， 6.8 and 1.9 percentage points in TR （True Rumor）-F1 score. The results combined with those from ablation experiments and analysis of GAN-generated embeddings show that RD-CSGU can effectively detect rumor posts on social media platforms.

Key words: rumor detection, comment sequence, heterogeneous propagation network, generated feature, propagation structure

中图分类号:

TP391.1

孟文凡, 周丽华, 王晓旭. 融合评论序列二义性与生成用户隐私特征的谣言检测[J]. 计算机应用, 2024, 44(8): 2342-2350.

Wenfan MENG, Lihua ZHOU, Xiaoxu WANG. Rumor detection by fusing ambiguity in comment sequences and generating user privacy features[J]. Journal of Computer Applications, 2024, 44(8): 2342-2350.

图/表 8

参考文献 27

1	徐铭达，张子柯，许小可. 基于模体度的社交网络虚假信息传播机制研究［J］. 计算机研究与发展， 2021， 58（7）： 1425-1435.
	XU M D， ZHANG Z K， XU X K. Research on spreading mechanism of false information in social networks by motif degree ［J］. Journal of Computer Research and Development， 2021， 58（7）： 1425-1435.
2	YANG F， LIU Y， YU X， et al. Automatic detection of rumor on Sina Weibo［C］// Proceedings of the 2012 ACM SIGKDD Workshop on Mining Data Semantics. New York： ACM， 2012： 1-7.
3	KWON S， CHA M， JUNG K， et al. Prominent features of rumor propagation in online social media［C］// Proceedings of the 2013 IEEE 13th International Conference on Data Mining. Piscataway： IEEE， 2013：1103-1108.
4	CHEN T， LI X， YIN H， et al. Call attention to rumors： deep attention based recurrent neural networks for early rumor detection［C］// Proceedings of the 22th Pacific-Asia Conference on Knowledge Discovery and Data Mining， Cham： Springer， 2018： 40-52.
5	KWON S， CHA M， JUNG K. Rumor detection over varying time windows［J］. PLoS ONE， 2017， 12（1）： e0168344.
6	PENG Y， WANG J. Rumor detection based on attention CNN and time series of context information［J］. Future Internet， 2021， 13（11）： 267.
7	HUANG Q， ZHOU C， WU J， et al. Deep structure learning for rumor detection on Twitter［C］// Proceedings of the 2019 International Joint Conference on Neural Networks. Piscataway： IEEE， 2019： 1-8.
8	YUAN C， MA Q， ZHOU W， et al. Jointly embedding the local and global relations of heterogeneous graph for rumor detection［C］//Proceedings of the 2019 IEEE International Conference on Data Mining. Piscataway： IEEE， 2019： 796-805.
9	V-H NGUYEN， SUGIYAMA K， NAKOV P， et al. FANG： leveraging social context for fake news detection using graph representation［C］// Proceedings of the 29th ACM International Conference on Information & Knowledge Management. New York： ACM， 2020： 1165-1174.
10	周丽华，王家龙，王丽珍，等. 异质信息网络表征学习综述［J］. 计算机学报， 2022， 45（1）： 160-189.
	ZHOU L H， WANG J L， WANG L Z， et al. Heterogeneous information network representation learning： a survey ［J］. Chinese Journal of Computers， 2022， 45（1）： 160-189.
11	蒋宗礼，樊珂，张津丽. 基于生成对抗网络和元路径的异质网络表示学习［J］. 计算机科学， 2022， 49（1）： 133-139.
	JIANG Z L， FAN K， ZHANG J L. Generative adversarial network and meta-path based heterogeneous network representation learning ［J］. Computer Science， 2022， 49（1）： 133-139.
12	YU F， LIU Q， WU S， et al. A convolutional approach for misinformation identification［C］// Proceedings of the 26th International Joint Conference on Artificial Intelligence. Menlo Park： AAAI Press， 2017： 3901-3907.
13	LOTFI S， MIRZAREZAEE M， HOSSEINZADEH M， et al. Detection of rumor conversations in Twitter using graph convolutional networks［J］. Applied Intelligence， 2021， 51（7）： 4774-4787.
14	DE SILVA N， DOU D. Semantic oppositeness assisted deep contextual modeling for automatic rumor detection in social networks［C］// Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics. Stroudsburg： ACL， 2021： 405-415.
15	SHU K， ZHOU X， WANG S， et al. The role of user profiles for fake news detection［C］// Proceedings of 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. New York： ACM， 2019： 436-439.
16	HAMDI T， SLIMI H， BOUNHAS I， et al. A hybrid approach for fake news detection in twitter based on user features and graph embedding ［C］// Proceedings of the 16th International Conference of Distributed Computing and Internet Technology. Cham： Springer， 2020： 266-280.
17	JIANG S， CHEN X， ZHANG L， et al. User-characteristic enhanced model for fake news detection in social media［C］// Proceedings of the 8th CCF International Conference of Natural Language Processing and Chinese Computing. Cham： Springer， 2019： 634-646.
18	LU Y-J， LI C-T. GCAN： graph-aware co-attention networks for explainable fake news detection on social media［C］// Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Stroudsburg： ACL， 2020： 505-514.
19	BAI N， MENG F， RUI X， et al. Rumor detection based on a source-replies conversation tree convolutional neural net［J］. Computing， 2022， 104（5）： 1155-1171.
20	MA J， GAO W， K-F WONG. Detect rumors in microblog posts using propagation structure via kernel learning［C］// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics. Stroudsburg： ACL， 2017： 708-717.
21	BIAN T， XIAO X， XU T， et al. Rumor detection on social media with bi-directional graph convolutional networks［C］// Proceedings of the 34th AAAI Conference on Artificial Intelligence. Menlo Park： AAAI， 2020： 549-556.
22	HUANG Q， YU J， WU J， et al. Heterogeneous graph attention networks for early detection of rumors on Twitter［C］// Proceedings of the 2020 International Joint Conference on Neural Networks. Piscataway： IEEE， 2020： 1-8.
23	ZHANG X， ZHANG T， ZHAO W， et al. Dual-attention graph convolutional network［C］// Proceedings of the 5th Asian Conference of Pattern Recognition. Cham： Springer， 2020： 238-251.
24	MA J， GAO W， MITRA P， et al. Detecting rumors from microblogs with recurrent neural networks［C］// Proceedings of the 25th International Joint Conference on Artificial Intelligence. Menlo Park： AAAI Press， 2016： 3818-3824.
25	LIU Y， WU Y-F. Early detection of fake news on social media through propagation path classification with recurrent and convolutional networks［C］// Proceedings of the 32nd AAAI Conference on Artificial Intelligence. Menlo Park：AAAI， 2018： 354-361.
26	KHOO L M S， CHIEU H L， QIAN Z， et al. Interpretable rumor detection in microblogs by attending to user interactions［C］// Proceedings of the 34th AAAI Conference on Artificial Intelligence. Menlo Park： AAAI， 2020： 8783-8790.
27	LI J， BAO P， SHEN H， et al. MiSTR： a multiview structural-temporal learning framework for rumor detection［J］. IEEE Transactions on Big Data， 2022， 8（4）： 1007-1019.

数据集	帖子数	用户数	评论数	NR	FR	UR	TR
Twitter15	1 490	276 663	331 612	374	370	374	372
Twitter16	818	173 487	204 820	205	205	203	205
Weibo	4 664	2 746 818	3 805 656	0	2 313	0	2 351

数据集	帖子数	用户数	评论数	NR	FR	UR	TR
Twitter15	1 490	276 663	331 612	374	370	374	372
Twitter16	818	173 487	204 820	205	205	203	205
Weibo	4 664	2 746 818	3 805 656	0	2 313	0	2 351

模型	Twitter15					Twitter16
模型	Acc	NR-F1	FR-F1	TR-F1	UR-F1	Acc	NR-F1	FR-F1	TR-F1	UR-F1
GRU	0.646	0.792	0.574	0.608	0.592	0.633	0.772	0.489	0.686	0.593
PPC	0.842	0.811	0.875	0.818	0.790	0.863	0.820	0.898	0.843	0.837
PLAN	0.799	0.754	0.521	0.836	0.799	0.816	0.761	0.853	0.870	0.774
PTK	0.667	0.619	0.669	0.772	0.645	0.863	0.820	0.898	0.843	0.837
Bi-GCN	0.836	0.816	0.870	0.866	0.786	0.839	0.838	0.855	0.844	0.819
GLAN	0.905	0.924	0.917	0.852	0.927	0.902	0.921	0.869	0.847	0.968
MiSTR	0.862	0.848	0.885	0.861	0.854	0.865	0.849	0.882	0.878	0.852
RD-CSGU	0.914	0.939	0.905	0.878	0.933	0.924	0.936	0.907	0.915	0.936

模型	Twitter15					Twitter16
模型	Acc	NR-F1	FR-F1	TR-F1	UR-F1	Acc	NR-F1	FR-F1	TR-F1	UR-F1
GRU	0.646	0.792	0.574	0.608	0.592	0.633	0.772	0.489	0.686	0.593
PPC	0.842	0.811	0.875	0.818	0.790	0.863	0.820	0.898	0.843	0.837
PLAN	0.799	0.754	0.521	0.836	0.799	0.816	0.761	0.853	0.870	0.774
PTK	0.667	0.619	0.669	0.772	0.645	0.863	0.820	0.898	0.843	0.837
Bi-GCN	0.836	0.816	0.870	0.866	0.786	0.839	0.838	0.855	0.844	0.819
GLAN	0.905	0.924	0.917	0.852	0.927	0.902	0.921	0.869	0.847	0.968
MiSTR	0.862	0.848	0.885	0.861	0.854	0.865	0.849	0.882	0.878	0.852
RD-CSGU	0.914	0.939	0.905	0.878	0.933	0.924	0.936	0.907	0.915	0.936

模型	Acc	非谣言分类			谣言分类
模型	Acc	FR-Precision	FR-Recall	FR-F1	TR-Precision	TR-Recall	TR-F1
GRU	0.910	0.952	0.864	0.906	0.876	0.956	0.914
PPC	0.921	0.949	0.889	0.918	0.896	0.962	0.923
PLAN	0.922	0.920	0.930	0.911	0.923	0.914	0.932
PTK	0.891	0.907	0.868	0.887	0.876	0.913	0.894
Bi-GCN	0.924	0.923	0.927	0.919	0.925	0.921	0.929
GLAN	0.946	0.949	0.943	0.946	0.943	0.948	0.945
MiSTR	0.947	0.921	0.977	0.948	0.976	0.917	0.946
RD-CSGU	0.964	0.957	0.971	0.964	0.971	0.957	0.964

融合评论序列二义性与生成用户隐私特征的谣言检测

Rumor detection by fusing ambiguity in comment sequences and generating user privacy features

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 27

相关文章 3

编辑推荐

Metrics

模型	参数量	Twitter15		Twitter16		Weibo
模型	参数量	时间/s	内存/GB	时间/s	内存/GB	时间/s	内存/GB
RD-CSGU	10 958 798	24.0	31.7	20.4	30.0	78.9	42.7
Bi-GCN	1 289 476	73.2	19.9	161.5	17.5	1 127.3	35.4
GLAN	6 393 798	5.6	3.4	3.6	2.3	7.8	4.7

[1]	薛海涛, 王莉, 杨延杰, 廉飚. 基于用户传播网络与消息内容融合的谣言检测模型[J]. 《计算机应用》唯一官方网站, 2021, 41(12): 3540-3545.
[2]	刘政, 卫志华, 张韧弦. 基于卷积神经网络的谣言检测[J]. 计算机应用, 2017, 37(11): 3053-3056.
[3]	杨文太, 梁刚, 谢凯, 杨进, 许春. 基于突发话题和领域专家的微博谣言检测方法[J]. 计算机应用, 2017, 37(10): 2799-2805.