Hyperbolic graph topic modeling network with continuously updated topic tree

Connectivity across documents often exhibits a hierarchical network structure. Hyperbolic Graph Neural Networks (HGNNs) have shown promise in preserving network hierarchy. However, they do not model the notion of topics, thus document representations lack semantic interpretability. On the other hand...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHANG, Ce, YING, Rex, LAUW, Hady Wirawan
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2023
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/8309
https://ink.library.smu.edu.sg/context/sis_research/article/9312/viewcontent/kdd23.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-9312
record_format dspace
spelling sg-smu-ink.sis_research-93122023-12-05T03:13:53Z Hyperbolic graph topic modeling network with continuously updated topic tree ZHANG, Ce YING, Rex LAUW, Hady Wirawan Connectivity across documents often exhibits a hierarchical network structure. Hyperbolic Graph Neural Networks (HGNNs) have shown promise in preserving network hierarchy. However, they do not model the notion of topics, thus document representations lack semantic interpretability. On the other hand, a corpus of documents usually has high variability in degrees of topic specificity. For example, some documents contain general content (e.g., sports), while others focus on specific themes (e.g., basketball and swimming). Topic models indeed model latent topics for semantic interpretability, but most assume a flat topic structure and ignore such semantic hierarchy. Given these two challenges, we propose a Hyperbolic Graph Topic Modeling Network to integrate both network hierarchy across linked documents and semantic hierarchy within texts into a unified HGNN framework. Specifically, we construct a two-layer document graph. Intra- and cross-layer encoding captures network hierarchy. We design a topic tree for text decoding to preserve semantic hierarchy and learn interpretable topics. Supervised and unsupervised experiments verify the effectiveness of our model. 2023-08-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/8309 info:doi/10.1145/3580305.3599384 https://ink.library.smu.edu.sg/context/sis_research/article/9312/viewcontent/kdd23.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Hyperbolic graph neural networks Text mining Topic modeling Graphics and Human Computer Interfaces OS and Networks
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Hyperbolic graph neural networks
Text mining
Topic modeling
Graphics and Human Computer Interfaces
OS and Networks
spellingShingle Hyperbolic graph neural networks
Text mining
Topic modeling
Graphics and Human Computer Interfaces
OS and Networks
ZHANG, Ce
YING, Rex
LAUW, Hady Wirawan
Hyperbolic graph topic modeling network with continuously updated topic tree
description Connectivity across documents often exhibits a hierarchical network structure. Hyperbolic Graph Neural Networks (HGNNs) have shown promise in preserving network hierarchy. However, they do not model the notion of topics, thus document representations lack semantic interpretability. On the other hand, a corpus of documents usually has high variability in degrees of topic specificity. For example, some documents contain general content (e.g., sports), while others focus on specific themes (e.g., basketball and swimming). Topic models indeed model latent topics for semantic interpretability, but most assume a flat topic structure and ignore such semantic hierarchy. Given these two challenges, we propose a Hyperbolic Graph Topic Modeling Network to integrate both network hierarchy across linked documents and semantic hierarchy within texts into a unified HGNN framework. Specifically, we construct a two-layer document graph. Intra- and cross-layer encoding captures network hierarchy. We design a topic tree for text decoding to preserve semantic hierarchy and learn interpretable topics. Supervised and unsupervised experiments verify the effectiveness of our model.
format text
author ZHANG, Ce
YING, Rex
LAUW, Hady Wirawan
author_facet ZHANG, Ce
YING, Rex
LAUW, Hady Wirawan
author_sort ZHANG, Ce
title Hyperbolic graph topic modeling network with continuously updated topic tree
title_short Hyperbolic graph topic modeling network with continuously updated topic tree
title_full Hyperbolic graph topic modeling network with continuously updated topic tree
title_fullStr Hyperbolic graph topic modeling network with continuously updated topic tree
title_full_unstemmed Hyperbolic graph topic modeling network with continuously updated topic tree
title_sort hyperbolic graph topic modeling network with continuously updated topic tree
publisher Institutional Knowledge at Singapore Management University
publishDate 2023
url https://ink.library.smu.edu.sg/sis_research/8309
https://ink.library.smu.edu.sg/context/sis_research/article/9312/viewcontent/kdd23.pdf
_version_ 1784855628910428160