计算机科学

计算机科学 ›› 2026, Vol. 53 ›› Issue (8): 50-60.doi: 10.11896/jsjkx.250500069

• 数据库 & 大数据 & 数据科学 • 上一篇    下一篇

基于隐藏特征的增强GNNs跨网络身份关联算法

潘语泉1, 袁得嵛1,2, 王安然1, 贾源1   

  1. 1 中国人民公安大学信息网络安全学院 北京 100038
    2 安全防范与风险评估公安部重点实验室 北京 102623
  • 收稿日期:2025-05-19 修回日期:2025-07-18 出版日期:2026-08-15 发布日期:2026-08-17
  • 通讯作者: 袁得嵛(yuandeyu@ppsuc.edu.cn)
  • 作者简介:(1870722711@qq.com)
  • 基金资助:
    公安部技术研究计划重点项目(2024JSZ01)

Enhanced GNNs Across Social Networks User Identity Linkage Algorithm Based on HiddenFeatures

PAN Yuquan1, YUAN Deyu1,2, WANG Anran1, JIA Yuan1   

  1. 1 School of Information Network Security, People’s Security University of China, Beijing 100038, China
    2 Key Laboratory of Security and Risk Assessment, Ministry of Public Security, Beijing 102623, China
  • Received:2025-05-19 Revised:2025-07-18 Published:2026-08-15 Online:2026-08-17
  • About author:PAN Yuquan,born in 2001,postgra-duate.His main research interest is social network analysis.
    YUAN Deyu,born in 1986,Ph.D,asso-ciate professor,master’s supervisor.His main research interests include network security and social network analysis.
  • Supported by:
    Key Project of the Ministry of Public Security’s Technical Research Program(2024JSZ01).

PDF (PC)

121

摘要： 跨网络身份关联能够判别来自不同社交网络的虚拟用户是否属于同一自然人。为应对真实数据集中正负样本分布的均衡程度对用户身份判别准确性的影响,提出了基于隐藏特征的增强GNNs跨网络身份关联算法。首先,为充分利用社交网络结构信息,挖掘了其中的隐藏特征,明确了度较大的节点和一阶邻居节点的重要作用;其次,设计了w-LINE算法,用于生成结构特征向量,通过与结构特征值向量进行拼接,得到用户特征向量;再次,提出了增强GNNs,其由Feature-GCN,Dynamic-Weight-GAT,GAE-VAE组成,用于优化用户向量表达;最后,引入了自适应胶囊网络,自定义路由层和损失权重的设置,能够灵活应对不同的正负样本分布,实现跨网络身份关联。在两个真实数据集上开展对比实验,结果表明,与基线模型相比,所提算法在P,R,F1值中均有10%以上的提升。消融实验结果表明,各组件对算法的整体性能具有重要作用。

关键词: 跨社交网络, 身份关联, 图神经网络, 胶囊网络, 深度学习

Abstract: Across social networks user identification can determine whether virtual users from different social networks correspond to the same natural person.In order to address the influence of the equilibrium degree of the distribution of positive and negative samples in real datasets on the accuracy of user identity discrimination,an enhanced GNNs cross-network identity association algorithm based on hidden features is proposed.Firstly,in order to make full use of the structural information of social networks,the hidden features therein are mined,and it is clarified that nodes with larger degrees and first-order neighbor nodes play important roles.Secondly,the w-LINE algorithm is designed to generate the structural feature vectors.By concatenating them with the structural feature value vectors,the user feature vectors are obtained.Then,enhanced GNNs are proposed,including Feature-GCN,Dynamic-Weight-GAT,and GAE-VAE,for optimizing the expression of user vectors.Finally,the adaptive capsule network is introduced.With the settings of the custom routing layer and loss weights,it can flexibly cope with different positive and negative sample distributions and achieve cross-network identity association.Experiments are conducted on two real datasets.The comparative experimental results with the baseline model show that the proposed algorithm improves by more than 10% in the values of P,R,and F1.The results of the ablation experiment show that each component plays an important role in the overall performance of the proposed algorithm.

Key words: Across social networks, User identity linkage, Graph neural network, Capsule network, Deep learning

中图分类号: 

  • TP391
[1] WEI W,HUANG C,XIA L,et al.Contrastive meta learningwith behavior multiplicity for recommendation[C]//Procee-dings of the Fifteenth ACM International Conference on Web Search and Data Mining.2022:1120-1128.
[2] LI N,TSIGKANOS C,JIN Z,et al.Early validation of cyber-physical space systems via multi-concerns integration[J].Journal of Systems and Software,2020,170:110742.
[3] GAN Y,ZHANG C F,YANG R S.User identity alignmentacross heterogeneous networks based on meta-path attention[C]//International Conference on Computer Application and Information Security( ICCAIS 2021).2022.
[4] LI Y,LIU Q.A comprehensive review study ofcyber-attacks and cyber security;Emerging trends and recent developments[J].Energy Reports,2021,7:8176-8186.
[5] CHEN L,CHEN J,XIA C.Social network behavior and public opinionmanipulation[J].Journal of Information Security and Applications,2022,64:103060.
[6] LIU J L,LIU Y,MA C X,et al.User Identity Parsing in Heterogeneous Social Platforms[J].Data Collection and Processing,2022,37(5):1101-1114.
[7] ZHANG J,GUO Y G.Social Network User Identification Me-thod based on Preference Logic[J].Computer Simulation,2022,39(4):450-453,505.
[8] LIU Z L,QIN T,GUAN X H,et al.Online user identity attri-bute association method using user name similarity propagation model[J].Journal of Xi’an Jiaotong University,2016,50(4):1-6,27.
[9] ANISA H,AYDAY E.Profile matching across online social net-works[C]//Information and Communications Security:22nd International Conference.Springer,2020.
[10] LI Y,PENG Y,JI W,et al.User identificatio n based on display names across online social networks[J].IEEE Access,2017,5:17342-17353.
[11] DING X,ZHANG H,MA C,et al.User identification across multiple social networks based on naive Bayes model[J].IEEE Transactions on Neural Networks and Learning Systems,2024,35:4274-4285.
[12] QU Y T,XING L,MA H Z,et al.Exploiting user friendship networks for user identification across social networks[J].Symmetry,2022,14(1):110.
[13] LI Y J,PENG Y,JI W L,et al.User identification based on display names across online social networks[J].IEEE Access,2017,5:17342-17353.
[14] DAI J,MA Q.Cross-social network user matching based on user Checkin[J].Computer Engineering and Applications,2023,59(2):76-84.
[15] HUO T F.Cross-social media user identity linkingbased on deep modeling of user behavior[D].Beijing:University of Chinese Academy of Sciences,2022.
[16] Ll Y J,Jl W L,GAO X,et al.Matching user accounts with sp-atio-temporal awarenessacross social networks[J].Information Sciences,2021,570:1-15.
[17] ZHENG C H,LI P,PENG W.JORA:Weakly supervised useridentity linkage via jointly learning to represent and align[J].IEEE Transactions on Neural Networks and Learning Systems,2024,35(3):3900-3911.
[18] SENETTE C,MARCO S,MAURIZIO T.User Identity Linkage on Social Networks:A Review of Modern Techniques and Applications[J].IEEE Access,2024,12:171241-171268.
[19] DING F X,MA X Q,YANG Y,et al.User identity linkage across location-based social networks with spatio-temporal check-in patterns[C]//2020 IEEE International Conference on Parallel & Distributed Processing with Applications,Big Data & Cloud Computing,Sustainable Computing & Communications,Social Computing & Networking(ISPA/BDCloud/SocialCom/SustainCom).IEEE,2020.
[20] HANG Z B,GU Q H,YUE T,et al.Identifying the same person across two similar social networks in a unified way:Globally and locally[J].Information Sciences,2017,394(C):53-67.
[21] KONG X,ZHANG J,YU P S.Inferring Anchor Links across Multiple Heterogeneous Social Networks[C]//The 22nd ACM International Conference on Information & Knowledge Management,2013:179-188.
[22] MA X,DING F,PENG K,et al.CP-link:Exploiting continuous spatio-temporal check-in patterns for user identity linkag[J].IEEE Transactions on Mobile Computing,2023,22(8):4594-4606.
[23] LIU L,CHEN P,LI X,et al.Wlalign:Weisfeiler-lehman relabe-ling for aligning users across networks via regularized representation learning[J].IEEE Transactions on Knowledge and Data Engineering,2023,36(1):445-458.
[24] CHEN H,YIN H,SUN X,et al.Multi-level Graph Convolutio-nal Networks for Cross-platform Anchor LinkPrediction[C]//ACM SIGKDD Conference on Knowledge Discovery and Data Mining.2020:1503-1511.
[25] CHU X,FAN X,YAO D,et al.Cross-network embedding for multi-network alignment[C]//The World Wide Web Confe-rence.2019:273-284.
[26] MAN T,SHEN H,LIU S,et al.Predict anchor links across social networks via an embedding approach[C]//IJCAI.2016:1823-1829.
[27] KELKHA M M,RAHGOZAR M,ASADPOUR M.DeepLink:A novel link prediction framework based on deep learning[J].Journal of Information Science 2021,47(5):642-657.
[28] CHEN W,WANG W Q,YIN H Z,et al.HFUL:a hybrid framework for user account linkage across location-aware social networks[J].The VLDB Journal,2023,32(1):1-22.
[29] ZHOU F,WEN Z J,ZHONG T,et al.Unsupervised user identity linkage via graph neural networks[C]//GLOBECOM 2020-2020 IEEE Global Communications Conference.IEEE,2020.
[30] XIONG X,XIE X,WU Y,et al.DSANE:A dual structure-aware network embedding approach for user identity linkage[C]//Proceedings of the IEEE 8th International Conference on Big Data Analytics(ICBDA).2023:193-198.
[31] LI S D,LU D N,LI Q,et al.MFLink:User identity linkage across online social networks via multimodal fusion and adversarial learning[J].IEEE Transactions on Emerging Topics in Computational Intelligence,2024,8(5):3716-3725.
[32] HUANG S,XIANG H,LENG C,et al.Cross-Social-NetworkUser Identification Based on Bidirectional GCN and MNF-UI Models[J].Electronics,2024,13:2351.
[33] ZHOU F,LIU L,ZHANG K P,et al.DeepLink:A Deep Lear-ning Approach for User Identity Linkage[C]//IEEE Conference on Computer Communications.2018:1313-1321.
[34] TANG W,SUN H,WANG J,et al.Identifying users across social media networks for interpretable fine-grained neighborhood matching by adaptive GAT[J].IEEE Transactionas on Services Computing,2023,16(5):3453-3466.
[35] FENG J,ZHANG M Y,WANG H D,et al.DPLink:User identity linkage via deep neural network from heterogeneous mobility data[C]//The Web Conference.2019:459-469.
[36] ZHANG J,CHEN B,WANG X M,et al.MEgo2Vec:Embedding matched ego networks foruser alignment across social networks.[C]//International Conference on 16 Journal of Frontiers of Computer Science and Technology Information and Knowledge Management.2018:327-336.
[37] SHAO J,WANG Y,GAO H,et al.AsyLink:user identity linkage from text to geo-location via sparse labeled data[J].Neurocomputing.2023,515:174-184.
[38] ZHANG J W,YU S P.PCT:Partial co-alignment of social networks[C]//The Web Conference.2016:749-759.
[39] ZHANG Y T,TANG J,YANG Z L,et al.COSNET:Connecting heterogeneous social networks with local andglobal consistency[C]//ACM SIGKDD Conference on Knowledge Discovery and Data Mining.2015:1485-1494.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市两江新区洪湖西路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