Leveraging Entity-Type Properties in the Relational Context for Knowledge Graph Embedding

Md Mostafizur RAHMAN; Atsuhiro TAKASU

doi:10.1587/transinf.2019DAP0007

Leveraging Entity-Type Properties in the Relational Context for Knowledge Graph Embedding

Md Mostafizur RAHMAN, Atsuhiro TAKASU

Full Text Views

0

Cite this

Summary :

Knowledge graph embedding aims to embed entities and relations of multi-relational data in low dimensional vector spaces. Knowledge graphs are useful for numerous artificial intelligence (AI) applications. However, they (KGs) are far from completeness and hence KG embedding models have quickly gained massive attention. Nevertheless, the state-of-the-art KG embedding models ignore the category specific projection of entities and the impact of entity types in relational aspect. For example, the entity “Washington” could belong to the person or location category depending on its appearance in a specific relation. In a KG, an entity usually holds many type properties. It leads us to a very interesting question: are all the type properties of an entity are meaningful for a specific relation? In this paper, we propose a KG embedding model TPRC that leverages entity-type properties in the relational context. To show the effectiveness of our model, we apply our idea to the TransE, TransR and TransD. Our approach outperforms other state-of-the-art approaches as TransE, TransD, DistMult and ComplEx. Another, important observation is: introducing entity type properties in the relational context can improve the performances of the original translation distance based models.

Publication: IEICE TRANSACTIONS on Information Vol.E103-D No.5 pp.958-968

Publication Date: 2020/05/01

Publicized: 2020/02/03

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2019DAP0007

Type of Manuscript: Special Section PAPER (Special Section on Data Engineering and Information Management)

Category

Authors

Md Mostafizur RAHMAN
The Graduate University for Advanced Studies (SOKENDAI),National Institute of Informatics
Atsuhiro TAKASU
The Graduate University for Advanced Studies (SOKENDAI),National Institute of Informatics

Keyword

knowledge representations, entity types, embedding models

Cite this

Copy

Md Mostafizur RAHMAN, Atsuhiro TAKASU, "Leveraging Entity-Type Properties in the Relational Context for Knowledge Graph Embedding" in IEICE TRANSACTIONS on Information, vol. E103-D, no. 5, pp. 958-968, May 2020, doi: 10.1587/transinf.2019DAP0007.
Abstract: Knowledge graph embedding aims to embed entities and relations of multi-relational data in low dimensional vector spaces. Knowledge graphs are useful for numerous artificial intelligence (AI) applications. However, they (KGs) are far from completeness and hence KG embedding models have quickly gained massive attention. Nevertheless, the state-of-the-art KG embedding models ignore the category specific projection of entities and the impact of entity types in relational aspect. For example, the entity “Washington” could belong to the person or location category depending on its appearance in a specific relation. In a KG, an entity usually holds many type properties. It leads us to a very interesting question: are all the type properties of an entity are meaningful for a specific relation? In this paper, we propose a KG embedding model TPRC that leverages entity-type properties in the relational context. To show the effectiveness of our model, we apply our idea to the TransE, TransR and TransD. Our approach outperforms other state-of-the-art approaches as TransE, TransD, DistMult and ComplEx. Another, important observation is: introducing entity type properties in the relational context can improve the performances of the original translation distance based models.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019DAP0007/_p

Copy

@ARTICLE{e103-d_5_958,
author={Md Mostafizur RAHMAN, Atsuhiro TAKASU, },
journal={IEICE TRANSACTIONS on Information},
title={Leveraging Entity-Type Properties in the Relational Context for Knowledge Graph Embedding},
year={2020},
volume={E103-D},
number={5},
pages={958-968},
abstract={Knowledge graph embedding aims to embed entities and relations of multi-relational data in low dimensional vector spaces. Knowledge graphs are useful for numerous artificial intelligence (AI) applications. However, they (KGs) are far from completeness and hence KG embedding models have quickly gained massive attention. Nevertheless, the state-of-the-art KG embedding models ignore the category specific projection of entities and the impact of entity types in relational aspect. For example, the entity “Washington” could belong to the person or location category depending on its appearance in a specific relation. In a KG, an entity usually holds many type properties. It leads us to a very interesting question: are all the type properties of an entity are meaningful for a specific relation? In this paper, we propose a KG embedding model TPRC that leverages entity-type properties in the relational context. To show the effectiveness of our model, we apply our idea to the TransE, TransR and TransD. Our approach outperforms other state-of-the-art approaches as TransE, TransD, DistMult and ComplEx. Another, important observation is: introducing entity type properties in the relational context can improve the performances of the original translation distance based models.},
keywords={},
doi={10.1587/transinf.2019DAP0007},
ISSN={1745-1361},
month={May},}

Copy

TY - JOUR
TI - Leveraging Entity-Type Properties in the Relational Context for Knowledge Graph Embedding
T2 - IEICE TRANSACTIONS on Information
SP - 958
EP - 968
AU - Md Mostafizur RAHMAN
AU - Atsuhiro TAKASU
PY - 2020
DO - 10.1587/transinf.2019DAP0007
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 2020
AB - Knowledge graph embedding aims to embed entities and relations of multi-relational data in low dimensional vector spaces. Knowledge graphs are useful for numerous artificial intelligence (AI) applications. However, they (KGs) are far from completeness and hence KG embedding models have quickly gained massive attention. Nevertheless, the state-of-the-art KG embedding models ignore the category specific projection of entities and the impact of entity types in relational aspect. For example, the entity “Washington” could belong to the person or location category depending on its appearance in a specific relation. In a KG, an entity usually holds many type properties. It leads us to a very interesting question: are all the type properties of an entity are meaningful for a specific relation? In this paper, we propose a KG embedding model TPRC that leverages entity-type properties in the relational context. To show the effectiveness of our model, we apply our idea to the TransE, TransR and TransD. Our approach outperforms other state-of-the-art approaches as TransE, TransD, DistMult and ComplEx. Another, important observation is: introducing entity type properties in the relational context can improve the performances of the original translation distance based models.
ER -