Discovering Latent Representations of Relations for Interacting Systems

Systems whose entities interact with each other are common. In many interacting systems, it is difficult to observe the relations between entities which is the key information for analyzing the system. In recent years, there has been increasing interest in discovering the relationships between entit...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Dohae Lee, Young Jin Oh, In-Kwon Lee
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Graph neural network relational inference unsupervised learning Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/08fc98bf9acd467796283a6379dd2f60
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:08fc98bf9acd467796283a6379dd2f60
record_format	dspace
spelling	oai:doaj.org-article:08fc98bf9acd467796283a6379dd2f602021-11-18T00:04:34ZDiscovering Latent Representations of Relations for Interacting Systems2169-353610.1109/ACCESS.2021.3125335https://doaj.org/article/08fc98bf9acd467796283a6379dd2f602021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9600885/https://doaj.org/toc/2169-3536Systems whose entities interact with each other are common. In many interacting systems, it is difficult to observe the relations between entities which is the key information for analyzing the system. In recent years, there has been increasing interest in discovering the relationships between entities using graph neural networks. However, existing approaches are difficult to apply if the number of relations is unknown or if the relations are complex. We propose the DiScovering Latent Relation (DSLR) model, which is flexibly applicable even if the number of relations is unknown or many types of relations exist. The flexibility of our DSLR model comes from the design concept of our encoder that represents the relation between entities in a latent space rather than a discrete variable and a decoder that can handle many types of relations. We performed the experiments on synthetic and real-world graph data with various relationships between entities, and compared the qualitative and quantitative results with other approaches. The experiments show that the proposed method is suitable for analyzing dynamic graphs with an unknown number of complex relations.Dohae LeeYoung Jin OhIn-Kwon LeeIEEEarticleGraph neural networkrelational inferenceunsupervised learningElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 149089-149099 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Graph neural network relational inference unsupervised learning Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Graph neural network relational inference unsupervised learning Electrical engineering. Electronics. Nuclear engineering TK1-9971 Dohae Lee Young Jin Oh In-Kwon Lee Discovering Latent Representations of Relations for Interacting Systems
description	Systems whose entities interact with each other are common. In many interacting systems, it is difficult to observe the relations between entities which is the key information for analyzing the system. In recent years, there has been increasing interest in discovering the relationships between entities using graph neural networks. However, existing approaches are difficult to apply if the number of relations is unknown or if the relations are complex. We propose the DiScovering Latent Relation (DSLR) model, which is flexibly applicable even if the number of relations is unknown or many types of relations exist. The flexibility of our DSLR model comes from the design concept of our encoder that represents the relation between entities in a latent space rather than a discrete variable and a decoder that can handle many types of relations. We performed the experiments on synthetic and real-world graph data with various relationships between entities, and compared the qualitative and quantitative results with other approaches. The experiments show that the proposed method is suitable for analyzing dynamic graphs with an unknown number of complex relations.
format	article
author	Dohae Lee Young Jin Oh In-Kwon Lee
author_facet	Dohae Lee Young Jin Oh In-Kwon Lee
author_sort	Dohae Lee
title	Discovering Latent Representations of Relations for Interacting Systems
title_short	Discovering Latent Representations of Relations for Interacting Systems
title_full	Discovering Latent Representations of Relations for Interacting Systems
title_fullStr	Discovering Latent Representations of Relations for Interacting Systems
title_full_unstemmed	Discovering Latent Representations of Relations for Interacting Systems
title_sort	discovering latent representations of relations for interacting systems
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/08fc98bf9acd467796283a6379dd2f60
work_keys_str_mv	AT dohaelee discoveringlatentrepresentationsofrelationsforinteractingsystems AT youngjinoh discoveringlatentrepresentationsofrelationsforinteractingsystems AT inkwonlee discoveringlatentrepresentationsofrelationsforinteractingsystems
_version_	1718425204840988672

Discovering Latent Representations of Relations for Interacting Systems

Ejemplares similares