Generative temporal link prediction via self-tokenized sequence modeling

Wang, Y; Zhang, C; Wang, S; Yu, PS; Bai, L; Cui, L; Xu, G

Generative temporal link prediction via self-tokenized sequence modeling

Wang, Y Zhang, C Wang, S Yu, PS Bai, L Cui, L Xu, G

Permalink

Publisher:: Springer Science and Business Media LLC
Publication Type:: Journal Article
Citation:: World Wide Web, 2020, 23, (4), pp. 2471-2488
Issue Date:: 2020-07-01

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

The embargo period expires on 29 May 2022

Adobe PDF

Download Accepted versionAdobe PDF (1.49 MB)

View on publisher's site

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Wang, Y
dc.contributor.author	Zhang, C
dc.contributor.author	Wang, S
dc.contributor.author	Yu, PS
dc.contributor.author	Bai, L
dc.contributor.author	Cui, L
dc.contributor.author	Xu, G https://orcid.org/0000-0003-4493-6663
dc.date.accessioned	2020-09-30T11:20:57Z
dc.date.available	2020-09-30T11:20:57Z
dc.date.issued	2020-07-01
dc.identifier.citation	World Wide Web, 2020, 23, (4), pp. 2471-2488
dc.identifier.issn	1386-145X
dc.identifier.issn	1573-1413
dc.identifier.uri	http://hdl.handle.net/10453/142938
dc.description.abstract	© 2020, Springer Science+Business Media, LLC, part of Springer Nature. We formalize networks with evolving structures as temporal networks and propose a generative link prediction model, Generative Link Sequence Modeling (GLSM), to predict future links for temporal networks. GLSM captures the temporal link formation patterns from the observed links with a sequence modeling framework and has the ability to generate the emerging links by inferring from the probability distribution on the potential future links. To avoid overfitting caused by treating each link as a unique token, we propose a self-tokenization mechanism to transform each raw link in the network to an abstract aggregation token automatically. The self-tokenization is seamlessly integrated into the sequence modeling framework, which allows the proposed GLSM model to have the generalization capability to discover link formation patterns beyond raw link sequences. We compare GLSM with the existing state-of-art methods on five real-world datasets. The experimental results demonstrate that GLSM obtains future positive links effectively in a generative fashion while achieving the best performance (2-10% improvements on AUC) among other alternatives.
dc.language	en
dc.publisher	Springer Science and Business Media LLC
dc.relation.ispartof	World Wide Web
dc.relation.hasversion	Accepted manuscript version	en
dc.relation.isbasedon	10.1007/s11280-020-00821-y
dc.rights	This is a post-peer-review, pre-copyedit version of an article published in World Wide Web. The final authenticated version is available online at: https://doi.org/10.1007/s11280-020-00821-y	en
dc.rights	info:eu-repo/semantics/embargoedAccess
dc.subject	0804 Data Format, 0805 Distributed Computing, 0806 Information Systems
dc.subject.classification	Information Systems
dc.title	Generative temporal link prediction via self-tokenized sequence modeling
dc.type	Journal Article
utslib.citation.volume	23
utslib.for	0804 Data Format
utslib.for	0805 Distributed Computing
utslib.for	0806 Information Systems
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Strength - AAI - Advanced Analytics Institute Research Centre
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Computer Science
pubs.organisational-group	/University of Technology Sydney
utslib.copyright.status	open_access	*
utslib.copyright.embargo	2022-05-29T00:00:00+1000Z
dc.date.updated	2020-09-30T11:20:47Z
pubs.issue	4
pubs.publication-status	Published
pubs.volume	23
utslib.citation.issue	4

Abstract:

© 2020, Springer Science+Business Media, LLC, part of Springer Nature. We formalize networks with evolving structures as temporal networks and propose a generative link prediction model, Generative Link Sequence Modeling (GLSM), to predict future links for temporal networks. GLSM captures the temporal link formation patterns from the observed links with a sequence modeling framework and has the ability to generate the emerging links by inferring from the probability distribution on the potential future links. To avoid overfitting caused by treating each link as a unique token, we propose a self-tokenization mechanism to transform each raw link in the network to an abstract aggregation token automatically. The self-tokenization is seamlessly integrated into the sequence modeling framework, which allows the proposed GLSM model to have the generalization capability to discover link formation patterns beyond raw link sequences. We compare GLSM with the existing state-of-art methods on five real-world datasets. The experimental results demonstrate that GLSM obtains future positive links effectively in a generative fashion while achieving the best performance (2-10% improvements on AUC) among other alternatives.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/142938