计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (4): 314-323.doi: 10.11896/jsjkx.230200020

• 人工智能 • 上一篇    下一篇

基于残差网络融合多关系评论特征的虚假评论检测

雒泽阳1, 田华1, 窦英通2, 李曼文1, 张泽华1   

  1. 1 太原理工大学信息与计算机学院 山西 晋中030600
    2 伊利诺伊大学芝加哥分校 芝加哥60607
  • 收稿日期:2023-02-05 修回日期:2023-06-26 出版日期:2024-04-15 发布日期:2024-04-10
  • 通讯作者: 张泽华(zhangzehua@tyut.edu.cn)
  • 作者简介:(luozy5025@163.com)
  • 基金资助:
    国家自然科学基金(61702356,51901152);教育部产学合作协同育人项目(2020021680113);山西省回国留学人员科研资助项目(2020-040)

Fake Review Detection Based on Residual Networks Fusion of Multi-relationship Review Features

LUO Zeyang1, TIAN Hua1, DOU Yingtong2, LI Manwen1, ZHANG Zehua1   

  1. 1 College of Information and Computer,Taiyuan University of Technology,Jinzhong,Shanxi 030600,China
    2 University of Illinois at Chicago,Chicago 60607,USA
  • Received:2023-02-05 Revised:2023-06-26 Online:2024-04-15 Published:2024-04-10
  • Supported by:
    National Natural Science Foundation of China(61702356,51901152),Industry University Cooperation Education Program of the Ministry of Education(2020021680113) and Shanxi Scholarship Council of China(2020-040).

PDF (PC)

44

摘要： 随着电子商务和短视频社区平台的兴起,涌现出的虚假评论严重影响了用户体验。甚至为了对抗平台检测,伪装的评论(Review Camouflage)更加难以辨别。当前基于图神经网络(Graph Neural Networks,GNNs)的虚假评论检测方法在深层训练过程中容易出现网络退化和梯度消失问题。同时评论伪装导致评论标记更加倾斜,从而影响GNNs检测模型的鲁棒性。针对以上问题,提出了一种基于残差网络的检测方法 MRDRN,可融合多关系评论特征进行虚假评论识别。首先,为了减缓网络退化,结合残差网络进行深层评论特征提取,并给出一种新的邻居混合采样策略,可根据评论之间的特征相似性进行低阶及高阶邻居混合采样,从而缓解评论标记不均衡的问题并学习更加丰富的评论特征。其次,提出了一种多关系评论特征融合策略,通过关系内评论网络拓扑与多关系间评论特征的整体融合,来减小评论伪装的影响。在3个真实数据集上进行实验,结果表明,MRDRN相比基准方法具有更高的检测能力和更强的鲁棒性。

关键词: 虚假评论检测, 图神经网络, 残差网络, 评论伪装, 多关系特征融合

Abstract: With the rise of e-commerce and short video community platforms,the emergence of fake reviews has seriously affected user experience.Even to combat platform detection,review camouflage makes it harder to distinguish between true and false.Current fake review detection methods based on graph neural networks(GNNs) are prone to network degradation and gradient disappearance during deep training.At the same time,review camouflage causes review markers to skew more,which affects the robustness of GNNs detection model.To solve the above problems,a detection method based on residual network(MRDRN) is proposed,which can fuse the features of multi-relationship reviews to identify fake reviews.Firstly,in order to slow down network degradation,the feature extraction of deep reviews is carried out by combining residual network.A new neighbor mixed sampling strategy is proposed,which can be used to conduct low-and high-order neighbor mixed sampling according to the feature similarity between reviews,so as to alleviate the problem of imbalanced review marks and learn more rich review features.Secondly,a multi-relationship review features fusion strategy is proposed,which reduces the impact of review masking by integrating intra relationship review network topology and inter relationship review features as a whole.Experimental results on three real datasets show that MRDRN has higher detection capability and stronger robustness than the standard method.

Key words: Fake review detection, Graph neural network, Residual network, Review camouflage, Multi-relationship features fusion

中图分类号: 

  • TP389.1
[1]MOHAWESH R,XU S,TRAN S N,et al.Fake reviews detection:A survey[J].IEEE Access,2021,9:65771-65802.
[2]YAO J,ZHENG Y,JIANG H.An ensemble model for fake online review detection based on data resampling,feature pruning,and parameter optimization[J].IEEE Access,2021,9:16914-16927.
[3]SATIA BUDHI G,CHIONG R,WANG Z L,et al.Using a hybrid content-based and behaviour-based featuring approach in a parallel environment to detect fake reviews[J].Electronic Commerce Research and Applications,2021,47:101048.
[4]YU S,REN J,LI S,et al.Graph learning for fake review detection[J].Frontiers in Artificial Intelligence,2022,5:922589.
[5]NOEKHAH S,BINTI SALIM N,ZAKARIA N H.Opinionspam detection:Using multi-iterative graph-based model[J].Information Processing & Management,2020,57(1):102140.
[6]DOU Y,LIU Z,SUN L,et al.Enhancing Graph Neural Network-based Fraud Detectors Against Camouflaged Fraudsters[C]//Proceedings of the 29th ACM International Conference on Information & Knowledge Management.2020:315-324.
[7]ZHANG W,SHENG Z,JIANG Y,et al.Evaluating deep graph neural networks[J].arXiv:2108.00955,2021.
[8]LI G H,MÜLLER M,QIAN G C,et al.DeepGCNs:Making GCNs Go as Deep as CNNs[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(6):6923-6939.
[9]WANG X D,WANG X,SUN Z X.Network Traffic Anomaly Detection Method Based on Multi-scale Memory Residual Network[J].Computer Science,2022,49(8):314-322.
[10]PATEL N A,PATEL R.A Survey on Fake Review Detection Using Machine Learning Techniques[C]//2018 4th Interna-tional Conference on Computing Communication and Automation(ICCCA).IEEE,2018:1-6.
[11]HAJEK P,BARUSHKA A,MUNK M.Fake consumer review detection using deep neural networks integrating word embeddings and emotion mining[J].Neural Computing and Applications,2020,32:17259-17274.
[12]JIAN Y,CHEN X,WANG H.Fake Restaurant Review Detection Using Deep Neural Networks with Hybrid Feature Fusion Method[C]//27th International ConferenceDatabase Systems for Advanced Applications(DASFAA 2022).Virtual Event,Part III.Cham:Springer International Publishing,2022:133-148.
[13]WANG J,WEN R,WU C,et al.Fdgars:Fraudster Detection via Graph Convolutional Networks in Online App Review System[C]//Companion Proceedings of the 2019 World Wide Web Conference.2019:310-316.
[14]LI A,QIN Z,LIU R,et al.Spam Review Detection with Graph Convolutional Networks[C]//Proceedings of the 28th ACM International Conference on Information and Knowledge Management.2019:2703-2711.
[15]LIU Z,DOU Y,YU P S,et al.Alleviating the InconsistencyProblem of Applying Graph Neural Network to Fraud Detection[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval.2020:1569-1572.
[16]DOU Y.Robust Graph Learning for Misbehavior Detection[C]//Proceedings of the Fifteenth ACM International Confe-rence on Web Search and Data Mining.2022:1545-1546.
[17]LIU Y,AO X,ZHONG Q,et al.Alike and Unlike:Resolving Class Imbalance Problem in Financial Credit Risk Assessment[C]//Proceedings of the 29th ACM International Conference on Information & Knowledge Management.2020:2125-2128.
[18]CHI J,ZENG G,ZHONG Q,et al.Learning to Undersampling for Class Imbalanced Credit Risk Forecasting[C]//2020 IEEE International Conference on Data Mining(ICDM).IEEE,2020:72-81.
[19]LI X,SUN X,MENG Y,et al.Dice loss for data-imbalanced NLP tasks[J].arXiv:1911.02855,2019.
[20]SHU J,XIE Q,YI L,et al.Meta-weight-net:Learning an expli-cit mapping for sample weighting[J].arXiv:1902.07379,2019.
[21]SHI M,TANG Y,ZHU X,et al.Multi-class Imbalanced Graph Convolutional Network Learning[C]//Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence(IJCAI-20).2020.
[22]LIU Y,AO X,QIN Z,et al.Pick and Choose:A GNN-based Imbalanced Learning Approach for Fraud Detection[C]//Procee-dings of the Web Conference 2021.2021:3168-3177.
[23]ZENG Y,TANG J.Rlc-gnn:An improved deep architecture for spatial-based graph neural network with application to fraud detection[J].Applied Sciences,2021,11(12):5656.
[24]ZHANG Z,YI X,ZHAO X.Fake Speech Detection Using Residual Network with Transformer Encoder[C]//Proceedings of the 2021 ACM Workshop on Information Hiding and Multimedia Security.2021:13-22.
[25]YUNMEI S,BO Y,LE Z,et al.IMTS:Detecting fake reviews with image and text Semantics[J].Data Analysis and Know-ledge Discovery,2022,6(8):84-96.
[26]KIPF T N,WELLING M.Semi-supervised classification withgraph convolutional networks[J].arXiv:1609.02907,2016.
[27]VELICKOVIC P,CUCURULL G,CASANOVA A,et al.Graph attention networks[J].arXiv:1710.10903,2017.
[28]HAMILTON W L,YING R,LESKOVEC J.Inductive Representation Learning on Large Graphs[J].arXiv:1706.02216,2017.
[29]ZENG H,ZHOU H,SRIVASTAVA A,et al.Graphsaint:Graph sampling based inductive learning method[J].arXiv:1907.04931,2019.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发