"Knowledge Graph Embedding for Hyper-Relational Data" by Chunhong Zhang, Miao Zhou et al.

Home > TSINGHUA-SCIENCE-AND-TECHNOLOGY > Vol. 22 (2017) > Iss. 2

Article Title

Knowledge Graph Embedding for Hyper-Relational Data

Authors

Chunhong Zhang, the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China.
Miao Zhou, the Key Laboratory of Universal Wireless Communication, Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, China.
Xiao Han, the Key Laboratory of Universal Wireless Communication, Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, China.
Zheng Hu, the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China.
Yang Ji, the Key Laboratory of Universal Wireless Communication, Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, China.

Keywords

distributed representation, transfer matrix, knowledge graph embedding

Abstract

Knowledge graph representation has been a long standing goal of artificial intelligence. In this paper, we consider a method for knowledge graph embedding of hyper-relational data, which are commonly found in knowledge graphs. Previous models such as Trans (E, H, R) and CTransR are either insufficient for embedding hyper-relational data or focus on projecting an entity into multiple embeddings, which might not be effective for generalization nor accurately reflect real knowledge. To overcome these issues, we propose the novel model TransHR, which transforms the hyper-relations in a pair of entities into an individual vector, serving as a translation between them. We experimentally evaluate our model on two typical tasks—link prediction and triple classification. The results demonstrate that TransHR significantly outperforms Trans (E, H, R) and CTransR, especially for hyper-relational data.