一种基于深度学习的动态社交网络用户对齐方法

doi:10.12263/DZXB.20201436

电子学报 ›› 2022, Vol. 50 ›› Issue (8): 1925-1936.DOI: 10.12263/DZXB.20201436

一种基于深度学习的动态社交网络用户对齐方法

王飞扬, 冀鹏欣, 孙笠, 危倩, 李根, 张忠宝

北京邮电大学计算机学院，北京 100876

收稿日期:2020-12-15 修回日期:2021-04-11 出版日期:2022-08-25
- 通讯作者:
- 张忠宝
- 作者简介:
- 王飞扬男，1999年生，山东德州人.北京邮电大学博士研究生.2020年获得北邮计算机科学与技术学士学位.主要研究方向为机器学习、社交网络分析、图神经网络.E-mail: fywang@bupt.edu.cn
  冀鹏欣男，1995年生，河北邯郸人.2017年获得计算机科学与技术学士学位，2020年获得北邮计算机科学与技术硕士学位.主要研究方向为大数据、机器学习、社交网络分析.E-mail: jpx@bupt.edu.cn
  孙笠男，1994年生，河北唐山人.北京邮电大学博士研究生.2016年获得北邮物联网工程学士学位.主要研究方向为机器学习、社交网络分析、图神经网络、非欧几何机器学习.E-mail: l.sun@bupt.edu.cn
  危倩女，1997年生，湖北荆门人.北京邮电大学硕士研究生.2019年获得北邮通信工程学士学位.主要研究方向为大数据与智能信息处理.E-mail: wei_qian@bupt.edu.cn
  李根男，1991年生，山东菏泽人.2017年获得应用物理学士学位，2020年获计算机科学与技术硕士学位.主要研究方向为机器学习、社交网络分析、图神经网络.E-mail: genli@bupt.edu.cn
  张忠宝男，1985年生，山东菏泽人.博士.北京邮电大学网络与交换技术国家重点实验室副教授、博士生导师.主要研究方向为大数据、人工智能、社交网络分析.
- 基金资助:
- 国家自然科学基金联合培育基金 (U1936103)

A Deep Learning Based Dynamic Social Network Alignment Method

WANG Fei-yang, JI Peng-xin, SUN Li, WEI Qian, LI Gen, ZHANG Zhong-bao

Department of Computer Science，Beijing University of Posts and Telecommunications，Beijing 100876，China

Received:2020-12-15 Revised:2021-04-11 Online:2022-08-25 Published:2022-09-08
- Corresponding author:
- ZHANG Zhong-bao

摘要/Abstract

摘要：

社交网络对齐旨在从不同的社交网络中识别出属于同一自然人的社交账户.现有的相关研究大多着眼于静态社交网络的对齐上，然而，社交网络是动态发展的.本文观察到，这种动态性可以揭示出更多的决定性模式，从而更有利于社交网络的对齐，这种现象促使本文在动态场景中重新思考这个问题.于是，本文利用社交网络的动态性，设计一个深度学习架构来解决动态社交网络的对齐问题，其称为DeepDSA（Deep learning based Dynamic Social network Alignment method）.首先设计一个深度序列模型来分别捕捉社交网络结构和属性的动态性；其次，对于每一个社交网络，通过保持相同用户结构和属性之间的相关性来融合二元动态，得到原始的综合嵌入表示；最后，以半监督的方式进行空间变换学习，并将每个网络的原始嵌入投影到一个目标子空间中，在该子空间中自然人是唯一表示的.本文在真实世界的数据集上进行大量的实验，证明DeepDSA方法相较于目前的主流算法提升了10%的对齐效果.

关键词: 社交网络对齐, 动态性, 深度学习, 特征融合, 子空间学习

Abstract:

Social network alignment aims to identify social accounts belonging to the same natural person from different social networks. Most of the existing related researches focus on the alignment of static social networks. However, social networks are dynamically evolving. We observe that dynamics can reveal more discriminative patterns and thus can benefit social network alignment. This phenomenon motivates us to rethink this issue in dynamic scenarios. Therefore, we propose to leverage the dynamics of social networks and design a deep learning architecture to address the dynamic social network alignment problem, termed as DeepDSA. Specifically, we first design a deep sequence model to capture the dynamics of social network structure and attributes respectively. For each social network, we merged binary dynamics by maintaining the correlation between structure and attributes of the same user to obtain the original comprehensive embeddings. We finally perform spatial transformation learning in a semi-supervised manner, and project the original embedding of each network into a target subspace in which a natural person is uniquely represented. We conduct extensive experiments on real-world datasets and demonstrate the proposed DeepDSA achieves 10% improvement of precision against the current mainstream algorithm.

Key words: social network alignment? dynamics? deep learning? feature fusion? subspace learning

中图分类号:

TP311

王飞扬, 冀鹏欣, 孙笠, 危倩, 李根, 张忠宝. 一种基于深度学习的动态社交网络用户对齐方法[J]. 电子学报, 2022, 50(8): 1925-1936.

WANG Fei-yang, JI Peng-xin, SUN Li, WEI Qian, LI Gen, ZHANG Zhong-bao. A Deep Learning Based Dynamic Social Network Alignment Method[J]. Acta Electronica Sinica, 2022, 50(8): 1925-1936.

图/表 13

图1 利用结构和属性的动态社交网络对齐样例

表1 主要符号和定义

符号	定义
$G$	动态社交网络
$V$	用户集合
$S$	结构特征集合
$A$	属性特征集合
$P$	已知对齐结点集合
$U$	嵌入矩阵
$Q$	空间映射矩阵
$I$	指示矩阵
$M$	负指示矩阵
$β$ , $γ$	模型参数

表1 主要符号和定义

符号	定义
$G$	动态社交网络
$V$	用户集合
$S$	结构特征集合
$A$	属性特征集合
$P$	已知对齐结点集合
$U$	嵌入矩阵
$Q$	空间映射矩阵
$I$	指示矩阵
$M$	负指示矩阵
$β$ , $γ$	模型参数

图2 GRU细胞的内部结构

图3 DeepDSA模型图（a）输入（b）建模（c）对齐

图4 学习文本向量的框架

表2 数据集统计信息

数据集	节点	链接	文本	对齐数
Twitter	7 440	72 658	3 467 472	3 896
Foursquare	6 771	69 350	289 674	3 896
Online	34 737	1 294 814	155 296	33 158
Offline	34 076	1 228 485	101 648	33 158

图5 TF数据集不同k值下的实验结果

图6 豆瓣数据集不同k值下的实验结果

图7 TF数据集不同重叠率λ下的实验结果

图8 TF数据集不同训练率η下的实验结果

图9 TF数据集不同快照设置下的实验结果

图10 TF数据集不同维度下的实验结果

图11 豆瓣数据集结构、属性和时间信息对实验结果的影响

参考文献 31

1	CAO X Z, YU Y. BASS: A Bootstrapping approach for aligning heterogenous social networks[C]//Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Riva del Garda: Springer, 2016: 459-475.
2	ZHOU J Y, FAN J X. TransLink: User identity linkage across heterogeneous social networks via translating embeddings[C]//IEEE INFOCOM 2019-IEEE Conference on Computer Communications. Paris: IEEE, 2019: 2116-2124.
3	ZHOU F, LIU L, ZHANG K P, et al. DeepLink: A deep learning approach for user identity linkage[C]//IEEE INFOCOM 2018-IEEE Conference on Computer Communications. Honolulu: IEEE, 2018: 1313-1321.
4	LI C Z, WANG S Z, YU P S, et al. Distribution distance minimization for unsupervised user identity linkage[C]//CIKM'18: Proceedings of the 27th ACM International Conference on Information and Knowledge Management. Torino: ACM, 2018: 447-456.
5	LIU L, CHEUNG W K, LI X, et al. Aligning users across social networks using network embedding[C]//IJCAI'16: Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence. New York: AAAI Press, 2016. 1774-1780.
6	MAN T, SHEN H, LIU S, et al. Predict anchor links across social networks via an embedding approach[C]//IJCAI'16: Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence. New York: AAAI Press, 2016. 1823-1829.
7	SU S, SUN L, ZHANG Z B, et al. MASTER: Across multiple social networks, integrate attribute and structure embedding for reconciliation[C]//IJCAI'18: Proceedings of the 27th International Joint Conference on Artificial Intelligence. Stockholm: AAAI Press, 2018: 3863-3869.
8	TAN S L, GUAN Z Y, CAI D, et al. Mapping users across networks by manifold alignment on hypergraph[C]//AAAI'14: Proceedings of the Twenty-Eighth AAAI Conference on Artificial Intelligence. Québec City: AAAI Press, 2014: 159-165.
9	WANG Y Q, SHEN H W, GAO J H, et al. Learning binary hash codes for fast anchor link retrieval across networks[C]//WWW'19: The World Wide Web Conference. San Francisco: ACM, 2019: 3335-3341.
10	SHU K, WANG S H, TANG J L, et al. User identity linkage across online social networks[J]. ACM SIGKDD Explorations Newsletter, 2017, 18(2): 5-17.
11	ZHANG S, TONG H H, MACIEJEWSKI R, et al. Multilevel network alignment[C]//WWW'19: The World Wide Web Conference. San Francisco: ACM, 2019: 2344-2354.
12	XIE W, MU X, LEE R K W, et al. Unsupervised user identity linkage via factoid embedding[C]//2018 IEEE International Conference on Data Mining. Singapore: IEEE, 2018: 1338-1343.
13	ZHOU F, WEN Z J, TRAJCEVSKI G, et al. Disentangled network alignment with matching explainability[C]//IEEE INFOCOM 2019-IEEE Conference on Computer Communications. Paris: IEEE, 2019: 1360-1368.
14	ZAFARANI R, LIU H. Connecting corresponding identities across communities[C]//AAAI International Conference on Web and Social Media. Palo Alto: AAAI Press, 2009. 354-357.
15	ZHANG J, CHEN B, WANG X M, et al. MEgo2Vec: Embedding matched ego networks for user alignment across social networks[C]//CIKM'18: Proceedings of the 27th ACM International Conference on Information and Knowledge Management. Torino: ACM, 2018: 327-336.
16	ZHANG Y T, TANG J, YANG Z L, et al. COSNET: Connecting heterogeneous social networks with local and global consistency[C]//Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Sydney: ACM, 2015: 1485-1494.
17	ZHONG Z, CAO Y, GUO M, et al. Colink: An unsupervised framework for user identity linkage[C]//AAAI Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2018. 5714-5721.
18	ZAFARANI R, LIU H. Connecting users across social media sites: A behavioral-modeling approach[C]//Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Chicago: ACM, 2013: 41-49.
19	SUN L, ZHANG Z B, JI P X, et al. DNA: Dynamic social network alignment[C]//2019 IEEE International Conference on Big Data(Big Data). Los Angeles: IEEE, 2019: 1224-1231.
20	WANG Y Q, SHEN H W, LIU S H, et al. Cascade dynamics modeling with attention-based recurrent neural network[C]//IJCAI'17: Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence.Melbourne: AAAI Press, 2017: 2985-2991.
21	MU X, ZHU F D, LIM E P, et al. User identity linkage by latent user space modelling[C]//KDD'16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco: ACM, 2016: 1775-1784.
22	GREFF K, SRIVASTAVA R K, KOUTNIK J, et al. LSTM: A search space odyssey[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(10): 2222-2232.
23	PEROZZI B, AL-RFOU R, SKIENA S. DeepWalk: Online learning of social representations[C]//Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining. New York: ACM, 2014: 701-710.
24	YANG C, LIU Z, ZHAO D, et al. Network representation learning with rich text information[C]//International Joint Conference on Artificial Intelligence. Buenos Aires: AAAI Press, 2015. 2111-2117.
25	LE Q, MIKOLOV T. Distributed representations of sentences and documents[C]//International Conference on Machine Learning. Beijing: JMLR.org, 2014: 1188-1196.
26	UTZ S, TANIS M, VERMEULEN I. It is all about being popular: The effects of need for popularity on social network site use[J]. Cyberpsychology, Behavior, and Social Networking, 2012, 15(1): 37-42.
27	ZUO Y, LIU G N, LIN H, et al. Embedding temporal network via neighborhood formation[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. London: ACM, 2018: 2857-2866.
28	LIANG S S, ZHANG X L, REN Z C, et al. Dynamic embeddings for user profiling in twitter[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. London: ACM, 2018: 1764-1773.
29	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems. Long Beach: Curran Associates Inc., 2017: 5998-6008.
30	GAO H C, HUANG H. Deep attributed network embedding[C]//IJCAI'18: Proceedings of the 27th International Joint Conference on Artificial Intelligence. Stockholm: AAAI Press, 2018: 3364-3370.
31	KONG X N, ZHANG J W, YU P S. Inferring anchor links across multiple heterogeneous social networks[C]//CIKM'13: Proceedings of the 22nd ACM International Conference on Information & Knowledge Management. San Francisco: ACM, 2013: 179-188.

[1]	吴靖, 叶晓晶, 黄峰, 陈丽琼, 王志锋, 刘文犀. 基于深度学习的单帧图像超分辨率重建综述[J]. 电子学报, 2022, 50(9): 2265-2294.
[2]	李雪莹, 王田路, 梁鹏, 王翀. 基于系统模型的用户评论中非功能需求的自动分类[J]. 电子学报, 2022, 50(9): 2079-2089.
[3]	琚长瑞, 秦晓燕, 袁广林, 李豪, 朱虹. 尺度敏感损失与特征融合的快速小目标检测方法[J]. 电子学报, 2022, 50(9): 2119-2126.
[4]	张志昌, 于沛霖, 庞雅丽, 朱林, 曾扬扬. SMGN：用于对话状态跟踪的状态记忆图网络[J]. 电子学报, 2022, 50(8): 1851-1858.
[5]	张亚洲, 俞洋, 朱少林, 陈锐, 戎璐, 梁辉. 一种量子概率启发的对话讽刺识别网络模型[J]. 电子学报, 2022, 50(8): 1885-1893.
[6]	徐兴荣, 刘聪, 李婷, 郭娜, 任崇广, 曾庆田. 基于双向准循环神经网络和注意力机制的业务流程剩余时间预测方法[J]. 电子学报, 2022, 50(8): 1975-1984.
[7]	欧阳与点, 谢鲲, 谢高岗, 文吉刚. 面向大规模网络测量的数据恢复算法：基于关联学习的张量填充[J]. 电子学报, 2022, 50(7): 1653-1663.
[8]	裴炤, 邱文涛, 王淼, 马苗, 张艳宁. 基于Transformer动态场景信息生成对抗网络的行人轨迹预测方法[J]. 电子学报, 2022, 50(7): 1537-1547.
[9]	李超, 黄新宇, 王凯. 基于特征融合和自学习锚框的高分辨率图像小目标检测算法[J]. 电子学报, 2022, 50(7): 1684-1695.
[10]	彭闯, 王伦文, 胡炜林. 融合深度特征的电磁频谱异常检测算法[J]. 电子学报, 2022, 50(6): 1359-1369.
[11]	杨伟超, 杜宇, 文伟, 侯舒维, 徐常志, 张建华. 基于多重分形谱智能分析的卫星信号调制识别研究[J]. 电子学报, 2022, 50(6): 1336-1343.
[12]	张波, 陆云杰, 秦东明, 邹国建. 一种卷积自编码深度学习的空气污染多站点联合预测模型[J]. 电子学报, 2022, 50(6): 1410-1427.
[13]	李政伟, 李佳树, 尤著宏, 聂茹, 赵欢, 钟堂波. 基于异质图注意力网络的miRNA与疾病关联预测算法[J]. 电子学报, 2022, 50(6): 1428-1435.
[14]	廖勇, 李玉杰. 一种轻量化低复杂度的FDD大规模MIMO系统CSI反馈方法[J]. 电子学报, 2022, 50(5): 1211-1217.
[15]	冀振燕, 韩梦豪, 宋晓军, 冯其波. 面向激光光条图像修复的循环相似度映射网络[J]. 电子学报, 2022, 50(5): 1234-1242.

一种基于深度学习的动态社交网络用户对齐方法

A Deep Learning Based Dynamic Social Network Alignment Method

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 31

相关文章 15

编辑推荐

Metrics