基于多种提及关系的社交媒体用户位置推断

doi:10.11959/j.issn.1000-436X.2020229

通信学报 ›› 2020, Vol. 41 ›› Issue (12): 72-81.doi: 10.11959/j.issn.1000-436X.2020229

基于多种提及关系的社交媒体用户位置推断

乔亚琼¹^,², 罗向阳¹^,², 马江涛³, 李晨亮⁴, 张萌¹^,², 李瑞祥¹^,²

¹ 信息工程大学网络空间安全学院，河南郑州 450001
² 数学工程与先进计算国家重点实验室，河南郑州 450001
³ 郑州轻工业大学计算机与通信工程学院，河南郑州 450001
⁴ 武汉大学国家网络安全学院，湖北武汉 430075

修回日期:2020-07-20 出版日期:2020-12-25 发布日期:2020-12-01
作者简介:乔亚琼（1981- ），女，河南开封人，信息工程大学博士生，主要研究方向为数据挖掘和社交网络分析。
罗向阳（1978- ），男，湖北荆门人，博士，信息工程大学教授、博士生导师，主要研究方向为网络与信息安全。
马江涛（1981- ），男，河南开封人，博士，郑州轻工业大学讲师，主要研究方向为数据挖掘和社交网络分析。
李晨亮（1983- ），男，湖北武汉人，博士，武汉大学副教授，主要研究方向为数据挖掘、机器学习、社交网络分析、信息安全、信息检索、文本/网络挖掘和自然语言处理。
张萌（1996- ），女，河南偃师人，信息工程大学博士生，主要研究方向为数据挖掘和社交网络分析。
李瑞祥（1994- ），男，湖南衡阳人，信息工程大学博士生，主要研究方向为网络实体定位、数据分析和信息安全。
基金资助:
国家自然科学基金资助项目(U1804263);国家自然科学基金资助项目(U1636219);国家自然科学基金资助项目(61872287);国家自然科学基金资助项目(U1736214);国家重点研发计划基金资助项目(2016QY01W0105);国家重点研发计划基金资助项目(2016YFB0801303);中原英才计划-中原科技创新领军人才基金资助项目(1052020KJLJ0025);河南省科技创新人才计划基金资助项目(184200510018);河南省科技攻关基金资助项目(202102310237)

Social media user geolocalization based on multiple mention relationships

Yaqiong QIAO¹^,², Xiangyang LUO¹^,², Jiangtao MA³, Chenliang LI⁴, Meng ZHANG¹^,², Ruixiang LI¹^,²

¹ Information Engineering University, Zhengzhou 450001, China
² State Key Laboratory of Mathematical Engineering and Advanced Computing, Zhengzhou 450001, China
³ School of Computer and Communication Engineering, Zhengzhou University of Light Industry, Zhengzhou 450001, China
⁴ School of Cyber Science and Engineering, Wuhan University, Wuhan 430075, China

Revised:2020-07-20 Online:2020-12-25 Published:2020-12-01
Supported by:
The National Natural Science Foundation of China(U1804263);The National Natural Science Foundation of China(U1636219);The National Natural Science Foundation of China(61872287);The National Natural Science Foundation of China(U1736214);The National Key Research and Development Program of China(2016QY01W0105);The National Key Research and Development Program of China(2016YFB0801303);Zhongyuan Talents Program-Zhongyuan Science and Technology Innovation Leading Talent Project(1052020KJLJ0025);The Plan for Scientific Inno-vation Talent of Henan Province(184200510018);The Scientific and Technological Project of Henan Province(202102310237)

摘要/Abstract

摘要：

针对现有基于生成文本和社交关系的联合位置推断方法对社交媒体中异质数据间的位置关联性挖掘不充分的问题，提出了一种基于多种提及关系的社交媒体用户位置推断方法。首先，综合考虑社交媒体文本中用户之间的提及关系、用户对位置指示词的提及关系和用户对地理名词的提及关系，构建包含用户、位置指示词和地理名词3种节点的异质网络；其次，基于共同提及关系提出用户-词语-位置简化算法来构建用户-位置异质网络，将位置邻近的用户更为紧密地联系起来；再次，提出有偏的随机游走算法对图中节点采样以充分探索网络结构，缓解已知位置的稀疏性问题；最后，采用基于多层感知机的神经网络分类器对用户进行位置推断。在 GEOTEXT、TW-US 和TW-WORLD这3个代表性Twitter数据集上的实验结果表明，所提方法可显著提高用户位置推断准确率。

关键词: 社交媒体, 异质网络, 用户位置推断, 提及关系

Abstract:

Aiming at the problem that the existing joint user geolocalization methods based on social media text and social relationships do not sufficiently mine the location correlation between heterogeneous data in social media, a social media user geolocalization method based on multiple mention relationships was proposed.First, a heterogeneous network was constructed by comprehensively considering the mention relationship between users, the user's mention relationship with location indicative words, and the user's mention relationship with geographic nouns.Then, a network simplification strategy was proposed to construct a user-location heterogeneous network that connects users live nearby more closely based on the common mention relationship.After that, a biased random walk algorithm was proposed for the graph node sampling to fully explore the network structure and alleviate the sparsity problem of known locations.Finally, a neural network classifier based on a multilayer perceptron was used to infer the user's location.Experimental results on three representative Twitter data sets of GEOTEXT, TW-US and TW-WORLD show that the proposed method can significantly improve the user geolocalization accuracy.

Key words: social media, heterogeneous network, user geolocalization, mention relationship

中图分类号:

TP391

乔亚琼, 罗向阳, 马江涛, 李晨亮, 张萌, 李瑞祥. 基于多种提及关系的社交媒体用户位置推断[J]. 通信学报, 2020, 41(12): 72-81.

Yaqiong QIAO, Xiangyang LUO, Jiangtao MA, Chenliang LI, Meng ZHANG, Ruixiang LI. Social media user geolocalization based on multiple mention relationships[J]. Journal on Communications, 2020, 41(12): 72-81.

图/表 8

图1

图2

表1

词语在不同位置的分布（GEOTEXT数据集）"

位置标签	经度	纬度				地理名词
l₇	-111.88°	33.39°	$A r i z o n a$	softball	bubbles	josh	cracker	johnson	pissing
l₂₀	-97.30°	32.63°	$A u s t i n$	$D a l l a s$	coconut	fallon	badu	chorus	jamming
l₂₃	-96.08°	39.7°	$k a n s a s$	lambert	machines	insomnia	mayo	guts	slang
l₂₉	-87.63°	41.73°	$C h i c a g o$	$C h i$	lbs	peppers	greg	julius	dez
l₅₅	-82.29°	39.32°	$C o l u m b u s$	$n a s h v i l l e$	sangria	gummy	graphic	ditto	lauryn

表1

图3

表2

表3

表4

不同异质网络构建方法的位置推断结果（GEOTEXT数据集）"

方法	Acc@161	mean/km	median/km	覆盖率
GCN^[15]	60%	546	45	95.52%
GELP-MEW	38%	996	465	$99 . 86 %$
GELP-MW	55%	683	79	$99 . 86 %$
GELP-I	60%	553	48	$99 . 86 %$
GELP-UW	63%	499	$38$	98.28%
GELP	$64 %$	$494$	$38$	98.28%

表4

图4

参考文献 29

[1]	KAWANAKA S , MORIWAKI D . Uplift modeling for location-based online advertising[C]// Proceedings of the 3rd ACM SIGSPATIAL International Workshop on Location-based Recommendations,Geosocial Networks and Geoadvertising. New York:ACM Press, 2019: 1-4.
[2]	SHAHRAKI Z K , FATEMI A , MALAZI H T . Evidential fine-grained event localization using Twitter[J]. Information Processing ＆ Management, 2019,55(6): 102045.1-102045.20
[3]	ZHOU N , ZHAO W X , ZHANG X ,et al. A general multi-context embedding model for mining human trajectory data[J]. IEEE Transactions on Knowledge ＆ Data Engineering, 2016,28(8): 1945-1958.
[4]	张文静, 刘樵, 朱辉 ,等. 基于信息论方法的多等级位置隐私度量与保护[J]. 通信学报, 2019,40(12): 51-59.
	ZHANG W J , LIU Q , ZHU H ,et al. Evaluation and protection of mul-ti-level location privacy based on an information theoretic approach[J]. Journal on Communications, 2019,40(12): 51-59.
[5]	李维皓, 曹进, 李晖 ,等. 基于位置服务隐私自关联的隐私保护方案[J]. 通信学报, 2019,40(5): 57-66.
	LI W H , CAO J , LI H ,et al. Privacy self-correlation privacy-preserving scheme in LBS[J]. Journal on Communications, 2019,40(5): 57-66.
[6]	POULSTON A , STEVENSON M , BONTCHEVA K . Hyperlocal home location identification of Twitter profiles[C]// Proceedings of the 28th ACM Conference on Hypertext and Social Media. New York:ACM Press, 2017: 45-54.
[7]	EISENSTEIN J , O’CONNOR B , SMITH N A ,et al. A latent variable model for geographic lexical variation[C]// Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing. Stroudsburg:Association for Computational Linguistics, 2010: 1277-1287.
[8]	AHMED A , HONG L , SMOLA A J . Hierarchical geographical modeling of user locations from social media posts[C]// Proceedings of the 22nd International World Wide Web Conference. New York:ACM Press, 2013: 25-36.
[9]	WING B , BALDRIDGE J . Hierarchical discriminative classification for text-based geolocation[C]// Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. Stroudsburg:Association for Computational Linguistics, 2014: 336-348.
[10]	SCHULZ A , HADJAKOS A , PAULHEIM H ,et al. A multi-indicator approach for geolocalization of tweets[C]// Proceedings of the Seventh International Conference on Weblogs and Social Media. Palo Alto:AAAI Press, 2013: 573-582.
[11]	RAHIMI A , VU D , COHN T ,et al. Exploiting text and network context for geolocation of social media users[C]// The 2015 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies. Stroudsburg:Association for Computational Linguistics, 2015: 1362-1367.
[12]	RAHIMI A , COHN T , BALDWIN T . A neural model for user geolocation and lexical dialectology[C]// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics. Stroudsburg:Association for Computational Linguistics, 2017,2: 209-216.
[13]	JURGENS D , . That’s what friends are for:inferring location in online social media platforms based on social relationships[C]// Proceedings of the Seventh International Conference on Weblogs and Social Media. Palo Alto:AAAI Press, 2013: 273-282.
[14]	RAHIMI A , COHN T , BALDWIN T . Twitter user geolocation using a unified text and network prediction model[C]// Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing of the Asian Federation of Natural Language Processing. Stroudsburg:Association for Computational Linguistics, 2015: 630-636.
[15]	RAHIMI A , COHN T , BALDWIN T . Semi-supervised user geolocation via graph convolutional networks[C]// Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics. Stroudsburg:Association for Computational Linguistics, 2018,1(1): 2009-2019.
[16]	DO T H , NGUYEN D M , TSILIGIANNI E ,et al. Twitter user geolocation using deep multiview learning[C]// 2018 IEEE International Conference on Acoustics,Speech and Signal Processing. Piscataway:IEEE Press, 2018: 6304-6308.
[17]	ZHONG T , WANG T , WANG J ,et al. Multiple-aspect attentional graph neural networks for online social network user localization[J]. IEEE Access, 2020,8: 95223-95234.
[18]	ROLLER S , SPERIOSU M , RALLAPALLI S ,et al. Supervised text-based geolocation using language models on an adaptive grid[C]// Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning. Stroudsburg:Association for Computational Linguistics, 2012: 1500-1510.
[19]	HAN B , COOK P , BALDWIN T . Text-based twitter user geolocation prediction[J]. Journal of Artificial Intelligence Research, 2014,49(1): 451-500.
[20]	KRISHNAMURTHY R . Knowledge enabled location prediction of Twitter users[D]. Dayton:Wright State University, 2015.
[21]	CHENG Z , CAVERLEE J , LEE K . You are where you tweet:a content-based approach to geo-locating twitter users[C]// Proceedings of the 19th ACM Conference on Information and Knowledge Management. New York:ACM Press, 2010: 759-768.
[22]	ZHENG X , HAN J , SUN A . A survey of location prediction on Twitter[J]. IEEE Transactions on Knowledge ＆ Data Engineering, 2018,30(9): 1652-1671.
[23]	WING B , BALDRIDGE J . Simple supervised document geolocation with geodesic grids[C]// Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics:Human Language Technologies. Stroudsburg:Association for Computational Linguistics, 2011: 955-964.
[24]	GROVER A , LESKOVEC J . Node2Vec:scalable feature learning for networks[C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York:ACM Press, 2016: 855-864.
[25]	MIKOLOV T , SUTSKEVER I , CHEN K ,et al. Distributed representations of words and phrases and their compositionality[J]. Advances in Neural Information Processing Systems, 2013,26(2): 3111-3119.
[26]	NAIR V , HINTON G E . Rectified linear units improve restricted boltzmann machines vinod nair[C]// Proceedings of the 27th International Conference on Machine Learning. Madison:Omni Press, 2010, 807-814.
[27]	AL-RFOU R , ALAIN G , ALMAHAIRI A ,et al. Theano:a Python framework for fast computation of mathematical expressions[J]. arXiv e-prints,1605.02688, 2016: 1-19.
[28]	KINGMA D P , BA J . Adam:a method for stochastic optimization[J]. arXiv Preprint,arXiv:1412.6980, 2014.
[29]	HAN B , COOK P , BALDWIN T . Geolocation Prediction in Social Media Data by Finding Location Indicative Words[C]// Proceedings of the 24th International Conference on Computational Linguistics. Mumbai:The COLING 2012 Organizing Committee, 2012: 1045-1062.

数据集	推文数/个	用户提及数/个	用户总数/个	训练集用户数/个	测试集用户数/个	开发集用户数/个	每个用户推文数/条
GEOTEXT	378×10³	109×10³	9 475	5 685	1 895	1 895	39
TW-US	38×10⁶	3.63×10⁶	450×10³	430×10³	10×10³	10×10³	84
TW-WORLD	12×10⁶	16.8×10⁶	1.4×10⁶	1.38×10⁶	10×10³	10×10³	9

数据来源	方法		GEOTEXT			TW-US			TW-WORLD
数据来源	方法	Acc@161	mean/km	median/km	Acc@161/km	mean/km	median/km	Acc@161	mean/km	median
	HierLR+k-d tree^[9]	—	—	—	48%	686	191	31%	1 669	509
Text	(MLP + k-d tree)^[12]	38%	844	389	54%	554	120	34%	1 456	415
	LR^[11]	38%	880	397	50%	686	159	32%	1 724	450
Network	MADCEL-W^[14]	58%	586	60	54%	705	116	45%	2 525	279
	GCN-LP^[15]	58%	576	56	53%	563	126	45%	2 357	279
	MLP-TXT+NET^[15]	58%	554	58	66%	420	56	58%	1 030	53
	MADCEL-W-MLP^[12]	59%	578	61	61%	515	77	53%	1 280	104
Text+Network	GCN^[15]	60%	546	45	62%	485	71	54%	1 130	108
	MAGNN^[17]	62%	514	38	63%	452	67	55%	1 107	102
	GELP	64%	489	39	67%	430	44	59%	1 015	50

基于多种提及关系的社交媒体用户位置推断

Social media user geolocalization based on multiple mention relationships

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 29

相关文章 3

Metrics

推荐阅读 0

[1]	钱榕, 许建婷, 张克君, 董宏宇, 邢方远. 隐马尔可夫模型的异质网络链接预测方法研究[J]. 通信学报, 2022, 43(5): 214-225.
[2]	王树鹏,陈明,吴广君. 面向互联网的大规模重复图像检索技术研究[J]. 通信学报, 2014, 35(12): 196-202.
[3]	王树鹏，陈明，吴广君. 面向互联网的大规模重复图像检索技术研究[J]. 通信学报, 2014, 35(12): 23-202.