GEMSEC: Graph Embedding with Self Clustering

Benedek Rózemberczki; Ryan Davies; Rik Sarkar; Charles Sutton

doi:10.1145/3341161.3342890

GEMSEC: Graph Embedding with Self Clustering

Benedek Rózemberczki, Ryan Davies, Rik Sarkar, Charles Sutton

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

Abstract / Description of output

Modern graph embedding procedures can efficiently process graphs with millions of nodes. In this paper, we propose GEMSEC – a graph embedding algorithm which learns a clustering of the nodes simultaneously with computing their embedding. GEMSEC is a general extension of earlier work in the domain of sequence-based graph embedding. GEMSEC places nodes in an abstract feature space where the vertex features minimize the negative log-likelihood of preserving sampled vertex neighborhoods, and it incorporates known social network properties through a machine learning regularization.

We present two new social network datasets and show that by simultaneously considering the embedding and clustering problems with respect to social properties, GEMSEC extracts high-quality clusters competitive with or superior to other community detection algorithms. In experiments, the method is found to be computationally efficient and robust to the choice of hyperparameters.

Original language	English
Title of host publication	ASONAM '19: Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining
Editors	Francesca Spezzano, Wei Chen, Xiaokui Xiao
Publisher	Association for Computing Machinery (ACM)
Pages	65-72
Number of pages	8
ISBN (Print)	9781450368681
DOIs	https://doi.org/10.1145/3341161.3342890
Publication status	Published - 27 Aug 2019
Event	IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2019 - Marriot Downtown Hotel, Vancouver, Canada Duration: 27 Aug 2019 → 30 Aug 2019 Conference number: 11 http://asonam.cpsc.ucalgary.ca/2019/

Conference

Conference	IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2019
Abbreviated title	ASONAM 2019
Country/Territory	Canada
City	Vancouver
Period	27/08/19 → 30/08/19
Internet address	http://asonam.cpsc.ucalgary.ca/2019/

Keywords / Materials (for Non-textual outputs)

community detection
clustering
node embedding
network embedding
feature extraction

Access to Document

10.1145/3341161.3342890Licence: All Rights Reserved

GEMSEC Graph_ROZEMBERCZKI_DoA200619_AFVAccepted author manuscript, 369 KB

https://dl.acm.org/doi/abs/10.1145/3341161.3342890Licence: All Rights Reserved

Cite this

@inproceedings{262b3ddcb91442ce9b0e5c0ec982aebc,

title = "GEMSEC: Graph Embedding with Self Clustering",

abstract = "Modern graph embedding procedures can efficiently process graphs with millions of nodes. In this paper, we propose GEMSEC – a graph embedding algorithm which learns a clustering of the nodes simultaneously with computing their embedding. GEMSEC is a general extension of earlier work in the domain of sequence-based graph embedding. GEMSEC places nodes in an abstract feature space where the vertex features minimize the negative log-likelihood of preserving sampled vertex neighborhoods, and it incorporates known social network properties through a machine learning regularization. We present two new social network datasets and show that by simultaneously considering the embedding and clustering problems with respect to social properties, GEMSEC extracts high-quality clusters competitive with or superior to other community detection algorithms. In experiments, the method is found to be computationally efficient and robust to the choice of hyperparameters.",

keywords = "community detection, clustering, node embedding, network embedding, feature extraction",

author = "Benedek R{\'o}zemberczki and Ryan Davies and Rik Sarkar and Charles Sutton",

year = "2019",

month = aug,

day = "27",

doi = "10.1145/3341161.3342890",

language = "English",

isbn = "9781450368681",

pages = "65--72",

editor = "Spezzano, { Francesca} and Wei Chen and Xiaokui Xiao",

booktitle = "ASONAM '19: Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining",

publisher = "Association for Computing Machinery (ACM)",

address = "United States",

note = "IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2019, ASONAM 2019 ; Conference date: 27-08-2019 Through 30-08-2019",

url = "http://asonam.cpsc.ucalgary.ca/2019/",

}

Rózemberczki, B, Davies, R, Sarkar, R & Sutton, C 2019, GEMSEC: Graph Embedding with Self Clustering. in F Spezzano, W Chen & X Xiao (eds), ASONAM '19: Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. Association for Computing Machinery (ACM), pp. 65-72, IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2019, Vancouver, Canada, 27/08/19. https://doi.org/10.1145/3341161.3342890

GEMSEC: Graph Embedding with Self Clustering. / Rózemberczki, Benedek; Davies, Ryan; Sarkar, Rik et al.
ASONAM '19: Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. ed. / Francesca Spezzano; Wei Chen; Xiaokui Xiao. Association for Computing Machinery (ACM), 2019. p. 65-72.

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

TY - GEN

T1 - GEMSEC: Graph Embedding with Self Clustering

AU - Rózemberczki, Benedek

AU - Davies, Ryan

AU - Sarkar, Rik

AU - Sutton, Charles

N1 - Conference code: 11

PY - 2019/8/27

Y1 - 2019/8/27

N2 - Modern graph embedding procedures can efficiently process graphs with millions of nodes. In this paper, we propose GEMSEC – a graph embedding algorithm which learns a clustering of the nodes simultaneously with computing their embedding. GEMSEC is a general extension of earlier work in the domain of sequence-based graph embedding. GEMSEC places nodes in an abstract feature space where the vertex features minimize the negative log-likelihood of preserving sampled vertex neighborhoods, and it incorporates known social network properties through a machine learning regularization. We present two new social network datasets and show that by simultaneously considering the embedding and clustering problems with respect to social properties, GEMSEC extracts high-quality clusters competitive with or superior to other community detection algorithms. In experiments, the method is found to be computationally efficient and robust to the choice of hyperparameters.

AB - Modern graph embedding procedures can efficiently process graphs with millions of nodes. In this paper, we propose GEMSEC – a graph embedding algorithm which learns a clustering of the nodes simultaneously with computing their embedding. GEMSEC is a general extension of earlier work in the domain of sequence-based graph embedding. GEMSEC places nodes in an abstract feature space where the vertex features minimize the negative log-likelihood of preserving sampled vertex neighborhoods, and it incorporates known social network properties through a machine learning regularization. We present two new social network datasets and show that by simultaneously considering the embedding and clustering problems with respect to social properties, GEMSEC extracts high-quality clusters competitive with or superior to other community detection algorithms. In experiments, the method is found to be computationally efficient and robust to the choice of hyperparameters.

KW - community detection

KW - clustering

KW - node embedding

KW - network embedding

KW - feature extraction

U2 - 10.1145/3341161.3342890

DO - 10.1145/3341161.3342890

M3 - Conference contribution

SN - 9781450368681

SP - 65

EP - 72

BT - ASONAM '19: Proceedings of the 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining

A2 - Spezzano, Francesca

A2 - Chen, Wei

A2 - Xiao, Xiaokui

PB - Association for Computing Machinery (ACM)

T2 - IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2019

Y2 - 27 August 2019 through 30 August 2019

ER -

GEMSEC: Graph Embedding with Self Clustering

Abstract / Description of output

Conference

Keywords / Materials (for Non-textual outputs)

Access to Document

Fingerprint

Cite this