Knowledge Association with Hyperbolic Knowledge Graph Embeddings

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 5704–5716, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 5704 Knowledge Association with Hyperbolic Knowledge Graph Embeddings Zequn Sun1, Muhao Chen2,3, Wei Hu1, Chengming Wang1, Jian Dai4, Wei Zhang4 1State Key Laboratory for Novel Software Technology, Nanjing University, China 2Department of Computer and Information Science, University of Pennsylvania, USA 3Information Sciences Institute, University of Southern California, USA 4Alibaba Group, China zqsun.nju@gmail.com, muhaoche@usc.edu, whu@nju.edu.cn cmwang.nju@gmail.com, {yiding.dj,lantu.zw}@alibaba-inc.com Abstract Capturing associations for knowledge graphs (KGs) through entity alignment, entity type in- ference and other related tasks benefits NLP applications with comprehensive knowledge representations. Recent related methods built on Euclidean embeddings are challenged by the hierarchical structures and different scales of KGs. They also depend on high embedding dimensions to realize enough expressiveness. Differently, we explore with low-dimensional hyperbolic embeddings for knowledge associa- tion. We propose a hyperbolic relational graph neural network for KG embedding and cap- ture knowledge associations with a hyperbolic transformation. Extensive experiments on en- tity alignment and type inference demonstrate the effectiveness and efficiency of our method. 1 Introduction Knowledge graphs (KGs) have emerged as the driv- ing force of many NLP applications, e.g., KBQA (Hixon et al., 2015), dialogue generation (Moon et al., 2019) and narrative prediction (Chen et al., 2019). Different KGs are usually extracted from separate data sources or contributed by people with different expertise. Therefore, it is natural for these KGs to constitute complementary knowledge of the world that can be expressed in different languages, structures and levels of specificity (Lehmann et al., 2015; Speer et al., 2017). Associating multiple KGs via entity alignment (Chen et al., 2017) or type inference (Hao et al., 2019) particularly provides downstream applications with more comprehensive knowledge representations. Entity alignment and type inference seek to find two kinds of knowledge associations, i.e., sameAs and instanceOf, respectively. An example showing such associations is given in Figure 1. Specifically, entity alignment is to find equivalent entities from different entity-level KGs, such as United States c'1 c'2 c'3 c'4 c'5 e'1 e'4 e'3 e'2 Concept level of KG2 Entity level of KG2 c1 c2 c3 c4 c6 c5 e1 e4 e3 e2 Concept level of KG1 Entity level of KG1 Concept Entity Relation instanceOf sameAs Figure 1: Illustration of two kinds of knowledge asso- ciations (i.e., sameAs and instanceOf) in KGs. in DBpedia and United States of America in Wiki- data. Type inference, on the other hand, associates a specific entity with a concept describing its type in- formation, such as United States and Country. The main difference lies in whether such knowledge associations express the same level of specificity or not. Challenged by the diverse schemata, relational structures and granularities of knowledge repre- sentations in different KGs (Nikolov et al., 2009), traditional symbolic methods usually fall short of supporting heterogeneous knowledge association (Suchanek et al., 2011; Lacoste-Julien et al., 2013; Paulheim and Bizer, 2013). Recently, increasing efforts have been put into exploring embedding- based methods (Chen et al., 2017; …

Content truncated to 500 words