Collective geographical embedding for geolocating social network users

Fengjiao Wang; Chun Ta Lu; Yongzhi Qu; Philip S. Yu

doi:10.1007/978-3-319-57454-7_47

Collective geographical embedding for geolocating social network users

Fengjiao Wang, Chun Ta Lu, Yongzhi Qu, Philip S. Yu

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

14 Scopus citations

Abstract

Inferring the physical locations of social network users is one of the core tasks in many online services, such as targeted advertisement, recommending local events, and urban computing. In this paper, we introduce the Collective Geographical Embedding (CGE) algorithm to embed multiple information sources into a low dimensional space, such that the distance in the embedding space reflects the physical distance in the real world. To achieve this, we introduced an embedding method with a location affinity matrix as a constraint for heterogeneous user network. The experiments demonstrate that the proposed algorithm not only outperforms traditional user geolocation prediction algorithms by collectively extracting relations hidden in the heterogeneous user network, but also outperforms state-of-the-art embedding algorithms by appropriately casting geographical information of check-in.

Original language	English (US)
Title of host publication	Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings
Editors	Kyuseok Shim, Jae-Gil Lee, Longbing Cao, Xuemin Lin, Jinho Kim, Yang-Sae Moon
Publisher	Springer Verlag
Pages	599-611
Number of pages	13
ISBN (Print)	9783319574530
DOIs	https://doi.org/10.1007/978-3-319-57454-7_47
State	Published - 2017
Externally published	Yes
Event	21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017 - Jeju, Korea, Republic of Duration: May 23 2017 → May 26 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10234 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017
Country/Territory	Korea, Republic of
City	Jeju
Period	5/23/17 → 5/26/17

Bibliographical note

Funding Information:
This work is supported in part by NSF through grants IIS-1526499, and CNS-1626432, and NSFC 61672313. Yongzhi Qu would like to acknowledge national natural science foundation of China (NSFC 51505353).

Publisher Copyright:
© 2017, Springer International Publishing AG.

Keywords

Geolocation
Geometric regularization
Geometrical embedding

Access

10.1007/978-3-319-57454-7_47

OpenUrl availability

Full text

Cite this

Wang, F., Lu, C. T., Qu, Y., & Yu, P. S. (2017). Collective geographical embedding for geolocating social network users. In K. Shim, J.-G. Lee, L. Cao, X. Lin, J. Kim, & Y.-S. Moon (Eds.), Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings (pp. 599-611). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10234 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-57454-7_47

Collective geographical embedding for geolocating social network users. / Wang, Fengjiao; Lu, Chun Ta; Qu, Yongzhi et al.
Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings. ed. / Kyuseok Shim; Jae-Gil Lee; Longbing Cao; Xuemin Lin; Jinho Kim; Yang-Sae Moon. Springer Verlag, 2017. p. 599-611 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10234 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Wang, F, Lu, CT, Qu, Y & Yu, PS 2017, Collective geographical embedding for geolocating social network users. in K Shim, J-G Lee, L Cao, X Lin, J Kim & Y-S Moon (eds), Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10234 LNAI, Springer Verlag, pp. 599-611, 21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017, Jeju, Korea, Republic of, 5/23/17. https://doi.org/10.1007/978-3-319-57454-7_47

Wang F, Lu CT, Qu Y, Yu PS. Collective geographical embedding for geolocating social network users. In Shim K, Lee JG, Cao L, Lin X, Kim J, Moon YS, editors, Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings. Springer Verlag. 2017. p. 599-611. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-57454-7_47

Wang, Fengjiao ; Lu, Chun Ta ; Qu, Yongzhi et al. / Collective geographical embedding for geolocating social network users. Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings. editor / Kyuseok Shim ; Jae-Gil Lee ; Longbing Cao ; Xuemin Lin ; Jinho Kim ; Yang-Sae Moon. Springer Verlag, 2017. pp. 599-611 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{8cf3403e25974fa19edbc12e2b5da143,

title = "Collective geographical embedding for geolocating social network users",

abstract = "Inferring the physical locations of social network users is one of the core tasks in many online services, such as targeted advertisement, recommending local events, and urban computing. In this paper, we introduce the Collective Geographical Embedding (CGE) algorithm to embed multiple information sources into a low dimensional space, such that the distance in the embedding space reflects the physical distance in the real world. To achieve this, we introduced an embedding method with a location affinity matrix as a constraint for heterogeneous user network. The experiments demonstrate that the proposed algorithm not only outperforms traditional user geolocation prediction algorithms by collectively extracting relations hidden in the heterogeneous user network, but also outperforms state-of-the-art embedding algorithms by appropriately casting geographical information of check-in.",

keywords = "Geolocation, Geometric regularization, Geometrical embedding",

author = "Fengjiao Wang and Lu, {Chun Ta} and Yongzhi Qu and Yu, {Philip S.}",

note = "Funding Information: This work is supported in part by NSF through grants IIS-1526499, and CNS-1626432, and NSFC 61672313. Yongzhi Qu would like to acknowledge national natural science foundation of China (NSFC 51505353). Publisher Copyright: {\textcopyright} 2017, Springer International Publishing AG.; 21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017 ; Conference date: 23-05-2017 Through 26-05-2017",

year = "2017",

doi = "10.1007/978-3-319-57454-7_47",

language = "English (US)",

isbn = "9783319574530",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "599--611",

editor = "Kyuseok Shim and Jae-Gil Lee and Longbing Cao and Xuemin Lin and Jinho Kim and Yang-Sae Moon",

booktitle = "Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings",

}

TY - GEN

T1 - Collective geographical embedding for geolocating social network users

AU - Wang, Fengjiao

AU - Lu, Chun Ta

AU - Qu, Yongzhi

AU - Yu, Philip S.

N1 - Funding Information: This work is supported in part by NSF through grants IIS-1526499, and CNS-1626432, and NSFC 61672313. Yongzhi Qu would like to acknowledge national natural science foundation of China (NSFC 51505353). Publisher Copyright: © 2017, Springer International Publishing AG.

PY - 2017

Y1 - 2017

N2 - Inferring the physical locations of social network users is one of the core tasks in many online services, such as targeted advertisement, recommending local events, and urban computing. In this paper, we introduce the Collective Geographical Embedding (CGE) algorithm to embed multiple information sources into a low dimensional space, such that the distance in the embedding space reflects the physical distance in the real world. To achieve this, we introduced an embedding method with a location affinity matrix as a constraint for heterogeneous user network. The experiments demonstrate that the proposed algorithm not only outperforms traditional user geolocation prediction algorithms by collectively extracting relations hidden in the heterogeneous user network, but also outperforms state-of-the-art embedding algorithms by appropriately casting geographical information of check-in.

AB - Inferring the physical locations of social network users is one of the core tasks in many online services, such as targeted advertisement, recommending local events, and urban computing. In this paper, we introduce the Collective Geographical Embedding (CGE) algorithm to embed multiple information sources into a low dimensional space, such that the distance in the embedding space reflects the physical distance in the real world. To achieve this, we introduced an embedding method with a location affinity matrix as a constraint for heterogeneous user network. The experiments demonstrate that the proposed algorithm not only outperforms traditional user geolocation prediction algorithms by collectively extracting relations hidden in the heterogeneous user network, but also outperforms state-of-the-art embedding algorithms by appropriately casting geographical information of check-in.

KW - Geolocation

KW - Geometric regularization

KW - Geometrical embedding

UR - http://www.scopus.com/inward/record.url?scp=85018453071&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85018453071&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-57454-7_47

DO - 10.1007/978-3-319-57454-7_47

M3 - Conference contribution

AN - SCOPUS:85018453071

SN - 9783319574530

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 599

EP - 611

BT - Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings

A2 - Shim, Kyuseok

A2 - Lee, Jae-Gil

A2 - Cao, Longbing

A2 - Lin, Xuemin

A2 - Kim, Jinho

A2 - Moon, Yang-Sae

PB - Springer Verlag

T2 - 21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017

Y2 - 23 May 2017 through 26 May 2017

ER -

Collective geographical embedding for geolocating social network users

Abstract

Publication series

Other

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this