Modeling the Correlations of Relations for Knowledge Graph Embedding

doi:10.1007/s11390-018-1821-8

	Modeling the Correlations of Relations for Knowledge Graph Embedding
	Zhu, Ji-Zhao 1,2; Jia, Yan-Tao 2; Xu, Jun 2; Qiao, Jian-Zhong 1; Cheng, Xue-Qi 2
	2018-03-01
发表期刊	JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY
ISSN	1000-9000
卷号	33 期号:2 页码:323-334
摘要	Knowledge graph embedding, which maps the entities and relations into low-dimensional vector spaces, has demonstrated its effectiveness in many tasks such as link prediction and relation extraction. Typical methods include TransE, TransH, and TransR. All these methods map different relations into the vector space separately and the intrinsic correlations of these relations are ignored. It is obvious that there exist some correlations among relations because different relations may connect to a common entity. For example, the triples (Steve Jobs, PlaceOfBrith, California) and (Apple Inc., Location, California) share the same entity California as their tail entity. We analyze the embedded relation matrices learned by TransE/TransH/TransR, and find that the correlations of relations do exist and they are showed as low-rank structure over the embedded relation matrix. It is natural to ask whether we can leverage these correlations to learn better embeddings for the entities and relations in a knowledge graph. In this paper, we propose to learn the embedded relation matrix by decomposing it as a product of two low-dimensional matrices, for characterizing the low-rank structure. The proposed method, called TransCoRe (Translation-Based Method via Modeling the Correlations of Relations), learns the embeddings of entities and relations with translation-based framework. Experimental results based on the benchmark datasets of WordNet and Freebase demonstrate that our method outperforms the typical baselines on link prediction and triple classification tasks.
关键词	knowledge graph embedding low-rank matrix decomposition
DOI	10.1007/s11390-018-1821-8
收录类别	SCI
语种	英语
资助项目	National Basic Research 973 Program of China[2014CB340405] ; National Key Research and Development Program of China[2016YFB1000902] ; National Natural Science Foundation of China[61402442] ; National Natural Science Foundation of China[61272177] ; National Natural Science Foundation of China[61173008] ; National Natural Science Foundation of China[61232010] ; National Natural Science Foundation of China[61303244] ; National Natural Science Foundation of China[61572469] ; National Natural Science Foundation of China[91646120] ; National Natural Science Foundation of China[61572473]
WOS研究方向	Computer Science
WOS类目	Computer Science, Hardware & Architecture ; Computer Science, Software Engineering
WOS记录号	WOS:000428379000007
出版者	SCIENCE PRESS
引用统计	被引频次：17[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/5976
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Xu, Jun; Qiao, Jian-Zhong
作者单位	1.Northeastern Univ, Coll Comp Sci & Engn, Shenyang 110169, Liaoning, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Key Lab Network Data Sci & Technol, Beijing 110190, Peoples R China
推荐引用方式 GB/T 7714	Zhu, Ji-Zhao,Jia, Yan-Tao,Xu, Jun,et al. Modeling the Correlations of Relations for Knowledge Graph Embedding[J]. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,2018,33(2):323-334.
APA	Zhu, Ji-Zhao,Jia, Yan-Tao,Xu, Jun,Qiao, Jian-Zhong,&Cheng, Xue-Qi.(2018).Modeling the Correlations of Relations for Knowledge Graph Embedding.JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,33(2),323-334.
MLA	Zhu, Ji-Zhao,et al."Modeling the Correlations of Relations for Knowledge Graph Embedding".JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 33.2(2018):323-334.