A Translation-Based Knowledge Graph Embedding Preserving Logical Property of Relations

Proceedings of NAACL-HLT 2016, pages 907–916, San Diego, California, June 12-17, 2016. c⃝2016 Association for Computational Linguistics A Translation-Based Knowledge Graph Embedding Preserving Logical Property of Relations Hee-Geun Yoon, Hyun-Je Song, Seong-Bae Park, Se-Young Park School of Computer Science and Engineering Kyungpook National University Daegu, 41566, Korea {hkyoon, hjsong, sbpark, sypark}@sejong.knu.ac.kr Abstract This paper proposes a novel translation-based knowledge graph embedding that preserves the logical properties of relations such as tran- sitivity and symmetricity. The embedding space generated by existing translation-based embeddings do not represent transitive and symmetric relations precisely, because they ignore the role of entities in triples. Thus, we introduce a role-specific projection which maps an entity to distinct vectors according to its role in a triple. That is, a head entity is pro- jected onto an embedding space by a head pro- jection operator, and a tail entity is projected by a tail projection operator. This idea is ap- plied to TransE, TransR, and TransD to pro- duce lppTransE, lppTransR, and lppTransD, respectively. According to the experimental results on link prediction and triple classifica- tion, the proposed logical property preserving embeddings show the state-of-the-art perfor- mance at both tasks. These results prove that it is critical to preserve logical properties of relations while embedding knowledge graphs, and the proposed method does it effectively. 1 Introduction Representing knowledge as a graph is one of the most effective ways to utilize human knowledge with a machine, and various large-scale knowledge graphs such as Freebase (Bollacker et al., 2008) and Yago (Suchanek et al., 2007) are available these days. However, the sparsity of the graphs makes it difficult to utilize them in real world applications. In spite of their huge volume, the relations among entities in the graphs are insufficient, which results in very limited inference of the knowledge of the graphs. Therefore, it is of importance to resolve such sparsity of knowledge graphs. One of the most promising methods to complete knowledge graphs is to embed the graphs in a low- dimensional continuous vector space. This method learns a vector representation of a knowledge graph, and the plausibility of a certain knowledge within the graph is measured with algebraic operations in the vector space. Thus, new knowledge can be harvested from the space by finding knowledge in- stances with high plausibility. The translation-based model among various knowledge-embedding models shows the state-of- the-art performance of knowledge graph completion (Bordes et al., 2013; Wang et al., 2014; Lin et al., 2015; Ji et al., 2015). TransE (Bordes et al., 2013) is one of the well-known translation-based approaches to this problem. When a set of knowledge triples (h, r, t) composed of a relation (r) and two enti- ties (h and t) is given, it finds vector representations of h, t, and r by compelling the vector of t to be same with the sum of the vectors of h and r. While TransE embeds all relations in a single vector space, TransH (Wang et al., 2014) and TransR (Lin et al., …

Content truncated to 500 words