File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Source-Aware Embedding Training on Heterogeneous Information Networks

TitleSource-Aware Embedding Training on Heterogeneous Information Networks
Authors
KeywordsAdversarial learning
Distribution alignment
Graph neural network
Graph representation learning
Heterogeneous information network
Recommendation system
Issue Date11-Feb-2023
PublisherMassachusetts Institute of Technology Press
Citation
Data Intelligence, 2023, v. 5, n. 3, p. 611-635 How to Cite?
Abstract

Heterogeneous information networks (HINs) have been extensively applied to real-world tasks, such as recommendation systems, social networks, and citation networks. While existing HIN representation learning methods can effectively learn the semantic and structural features in the network, little awareness was given to the distribution discrepancy of subgraphs within a single HIN. However, we find that ignoring such distribution discrepancy among subgraphs from multiple sources would hinder the effectiveness of graph embedding learning algorithms. This motivates us to propose SUMSHINE (Scalable Unsupervised Multi-Source Heterogeneous Information Network Embedding) — a scalable unsupervised framework to align the embedding distributions among multiple sources of an HIN. Experimental results on real-world datasets in a variety of downstream tasks validate the performance of our method over the state-of-the-art heterogeneous information network embedding algorithms.


Persistent Identifierhttp://hdl.handle.net/10722/337585
ISSN
2023 Impact Factor: 1.3
2023 SCImago Journal Rankings: 0.754
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorChan, TH-
dc.contributor.authorWong, CH-
dc.contributor.authorShen, J-
dc.contributor.authorYin, G-
dc.date.accessioned2024-03-11T10:22:15Z-
dc.date.available2024-03-11T10:22:15Z-
dc.date.issued2023-02-11-
dc.identifier.citationData Intelligence, 2023, v. 5, n. 3, p. 611-635-
dc.identifier.issn2641-435X-
dc.identifier.urihttp://hdl.handle.net/10722/337585-
dc.description.abstract<p>Heterogeneous information networks (HINs) have been extensively applied to real-world tasks, such as recommendation systems, social networks, and citation networks. While existing HIN representation learning methods can effectively learn the semantic and structural features in the network, little awareness was given to the distribution discrepancy of subgraphs within a single HIN. However, we find that ignoring such distribution discrepancy among subgraphs from multiple sources would hinder the effectiveness of graph embedding learning algorithms. This motivates us to propose SUMSHINE (Scalable Unsupervised Multi-Source Heterogeneous Information Network Embedding) — a scalable unsupervised framework to align the embedding distributions among multiple sources of an HIN. Experimental results on real-world datasets in a variety of downstream tasks validate the performance of our method over the state-of-the-art heterogeneous information network embedding algorithms.</p>-
dc.languageeng-
dc.publisherMassachusetts Institute of Technology Press-
dc.relation.ispartofData Intelligence-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.subjectAdversarial learning-
dc.subjectDistribution alignment-
dc.subjectGraph neural network-
dc.subjectGraph representation learning-
dc.subjectHeterogeneous information network-
dc.subjectRecommendation system-
dc.titleSource-Aware Embedding Training on Heterogeneous Information Networks-
dc.typeArticle-
dc.identifier.doi10.1162/dint_a_00200-
dc.identifier.scopuseid_2-s2.0-85173548662-
dc.identifier.volume5-
dc.identifier.issue3-
dc.identifier.spage611-
dc.identifier.epage635-
dc.identifier.eissn2641-435X-
dc.identifier.isiWOS:001065216500004-
dc.identifier.issnl2641-435X-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats