Cross-Sentence N-ary Relation Extraction using Lower-Arity Universal Schemas

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 6225–6231, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 6225 Cross-Sentence N-ary Relation Extraction using Lower-Arity Universal Schemas Kosuke Akimoto1, Takuya Hiraoka1, Kunihiko Sadamasa1, Mathias Niepert2 1 Security Research Laboratories, NEC Corporation 2 NEC Laboratories Europe {k-akimoto@ab,t-hiraoka@ce,k-sadamasa@az}.jp.nec.com mathias.niepert@neclab.eu Abstract Most existing relation extraction approaches exclusively target binary relations, and n-ary relation extraction is relatively unexplored. Current state-of-the-art n-ary relation extrac- tion method is based on a supervised learning approach and, therefore, may suffer from the lack of sufficient relation labels. In this paper, we propose a novel approach to cross-sentence n-ary relation extraction based on universal schemas. To alleviate the sparsity problem and to leverage inherent decomposability of n-ary relations, we propose to learn relation representations of lower-arity facts that result from decomposing higher-arity facts. The pro- posed method computes a score of a new n- ary fact by aggregating scores of its decom- posed lower-arity facts. We conduct exper- iments with datasets for ternary relation ex- traction and empirically show that our method improves the n-ary relation extraction perfor- mance compared to previous methods. 1 Introduction Relation extraction is a core natural language pro- cessing task which is concerned with the extrac- tion of relations between entities from text. It has numerous applications ranging from question an- swering (Xu et al., 2016) to automated knowledge base construction (Dong et al., 2014). While the vast majority of existing research fo- cuses on extracting binary relations, there exists only few recent approaches to extract n-ary re- lations, that is, relations among n ≥ 2 entities (Li et al., 2015; Ernst et al., 2018). In n-ary re- lation extraction, relation mentions tend to span multiple sentences more frequently as n increases. Thus, Peng et al. (2017) recently extended the problem to cross-sentence n-ary relation extrac- tion in which n-ary relations are extracted from multiple sentences. As a motivating example, con- sider the following text from Wikipedia: “Revis started off the 2009 season matched up against some of football’s best wide receivers. In Week 1, he helped limit Houston Texans Pro-bowler An- dre Johnson to four receptions for 35 yards.” In this example, two sentences collectively describes that Andre Johnson is a player of the football team the Texans during 2009 season, and thus we need cross-sentence information to correctly extract this ternary interaction among the three entities, i.e. Player(Andre Johnson, Texans, 2009 season). Previous methods (Peng et al., 2017; Song et al., 2018) capture cross-sentence n-ary relation men- tions by representing texts with a document graph which consists of both intra- and cross-sentence links between words. With this graphical repre- sentation, they applied graph neural networks to predict ternary relations in the medical domain. However, these methods train the neural networks in a supervised manner using distant supervision (Mintz et al., 2009) and, therefore, may suffer from the lack of sufficient positive labels when a well-populated …

Content truncated to 500 words