Relational graph convolutional networks: a closer look

Thiviyan Thanapalasingam; Lucas van Berkel; Peter Bloem; Paul Groth

doi:10.7717/PEERJ-CS.1073

Relational graph convolutional networks: a closer look

Thiviyan Thanapalasingam^*, Lucas van Berkel, Peter Bloem, Paul Groth

^*Corresponding author for this work

Research output: Contribution to Journal › Article › Academic › peer-review

Abstract

In this article, we describe a reproduction of the Relational Graph Convolutional Network (RGCN). Using our reproduction, we explain the intuition behind the model. Our reproduction results empirically validate the correctness of our implementations using benchmark Knowledge Graph datasets on node classification and link prediction tasks. Our explanation provides a friendly understanding of the different components of the RGCN for both users and researchers extending the RGCN approach. Furthermore, we introduce two new configurations of the RGCN that are more parameter efficient. The code and datasets are available at https://github.com/thiviyanT/torch-rgcn.

Original language	English
Article number	e1073
Pages (from-to)	1-33
Number of pages	33
Journal	PeerJ Computer Science
Volume	8
DOIs	https://doi.org/10.7717/PEERJ-CS.1073
Publication status	Published - 2 Nov 2022

Bibliographical note

Funding Information:
This research was supported by the responsible data science track of the VSNU Digital Society program. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Publisher Copyright:
© Copyright 2022 Thanapalasingam et al.

Keywords

Graph convolutional network
Knowledge graphs
Link prediction
Node classification
Relational graphs
Representation learning

Access to Document

10.7717/PEERJ-CS.1073

Persistent URL (handle)

Copy link address

Cite this

@article{558587976cec46c785016a1c8c88d319,

title = "Relational graph convolutional networks: a closer look",

abstract = "In this article, we describe a reproduction of the Relational Graph Convolutional Network (RGCN). Using our reproduction, we explain the intuition behind the model. Our reproduction results empirically validate the correctness of our implementations using benchmark Knowledge Graph datasets on node classification and link prediction tasks. Our explanation provides a friendly understanding of the different components of the RGCN for both users and researchers extending the RGCN approach. Furthermore, we introduce two new configurations of the RGCN that are more parameter efficient. The code and datasets are available at https://github.com/thiviyanT/torch-rgcn.",

keywords = "Graph convolutional network, Knowledge graphs, Link prediction, Node classification, Relational graphs, Representation learning",

author = "Thiviyan Thanapalasingam and {van Berkel}, Lucas and Peter Bloem and Paul Groth",

note = "Funding Information: This research was supported by the responsible data science track of the VSNU Digital Society program. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Publisher Copyright: {\textcopyright} Copyright 2022 Thanapalasingam et al.",

year = "2022",

month = nov,

day = "2",

doi = "10.7717/PEERJ-CS.1073",

language = "English",

volume = "8",

pages = "1--33",

journal = "PeerJ Computer Science",

issn = "2376-5992",

publisher = "PeerJ",

}

TY - JOUR

T1 - Relational graph convolutional networks

T2 - a closer look

AU - Thanapalasingam, Thiviyan

AU - van Berkel, Lucas

AU - Bloem, Peter

AU - Groth, Paul

N1 - Funding Information: This research was supported by the responsible data science track of the VSNU Digital Society program. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Publisher Copyright: © Copyright 2022 Thanapalasingam et al.

PY - 2022/11/2

Y1 - 2022/11/2

N2 - In this article, we describe a reproduction of the Relational Graph Convolutional Network (RGCN). Using our reproduction, we explain the intuition behind the model. Our reproduction results empirically validate the correctness of our implementations using benchmark Knowledge Graph datasets on node classification and link prediction tasks. Our explanation provides a friendly understanding of the different components of the RGCN for both users and researchers extending the RGCN approach. Furthermore, we introduce two new configurations of the RGCN that are more parameter efficient. The code and datasets are available at https://github.com/thiviyanT/torch-rgcn.

AB - In this article, we describe a reproduction of the Relational Graph Convolutional Network (RGCN). Using our reproduction, we explain the intuition behind the model. Our reproduction results empirically validate the correctness of our implementations using benchmark Knowledge Graph datasets on node classification and link prediction tasks. Our explanation provides a friendly understanding of the different components of the RGCN for both users and researchers extending the RGCN approach. Furthermore, we introduce two new configurations of the RGCN that are more parameter efficient. The code and datasets are available at https://github.com/thiviyanT/torch-rgcn.

KW - Graph convolutional network

KW - Knowledge graphs

KW - Link prediction

KW - Node classification

KW - Relational graphs

KW - Representation learning

UR - http://www.scopus.com/inward/record.url?scp=85143869291&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85143869291&partnerID=8YFLogxK

U2 - 10.7717/PEERJ-CS.1073

DO - 10.7717/PEERJ-CS.1073

M3 - Article

AN - SCOPUS:85143869291

SN - 2376-5992

VL - 8

SP - 1

EP - 33

JO - PeerJ Computer Science

JF - PeerJ Computer Science

M1 - e1073

ER -

Relational graph convolutional networks: a closer look

Abstract

Bibliographical note

Keywords

Access to Document

Persistent URL (handle)

Other files and links

Fingerprint

Cite this