A Semantic Filter Based on Relations for Knowledge Graph Completion

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7920–7929 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 7920 A Semantic Filter Based on Relations for Knowledge Graph Completion Zongwei Liang, Junan Yang, Hui Liu, Keju Huang National University of Defense Technology, Hefei, China zwliang17@nudt.edu.cn,yangjunan@ustc.edu.cn, {christ592604,huangkeju}@163.com Abstract Knowledge graph embedding, representing en- tities and relations in the knowledge graphs with high-dimensional vectors, has made sig- nificant progress in link prediction. More re- searchers have explored the representational capabilities of models in recent years. That is, they investigate better representational mod- els to fit symmetry/antisymmetry and combi- nation relationships. The current embedding models are more inclined to utilize the identi- cal vector for the same entity in various triples to measure the matching performance. The ob- servation that measuring the rationality of spe- cific triples means comparing the matching de- gree of the specific attributes associated with the relations is well-known. Inspired by this fact, this paper designs Semantic Filter Based on Relations(SFBR) to extract the required at- tributes of the entities. Then the rationality of triples is compared under these extracted attributes through the traditional embedding models. The semantic filter module can be added to most geometric and tensor decompo- sition models with minimal additional mem- ory. Experiments on the benchmark datasets show that the semantic filter based on relations can suppress the impact of other attribute di- mensions and improve link prediction perfor- mance. The tensor decomposition models with SFBR have achieved state-of-the-art. 1 Introduction Knowledge Graphs (KGs) are collections of large- scale triples, such as Freebase(Bordes et al., 2013) , YAGO (Suchanek et al., 2008) and DBpedia(Auer et al., 2007). KGs play a crucial role in applica- tions such as question answering services, search engines, and medical care. Although there are bil- lions of triples in KGs, they are still incomplete. These incomplete knowledge bases will bring limi- tations to practical applications. Therefore, knowl- edge graph completion, known as link prediction, which automatically predicts missing links between entities based on given links, has recently attracted growing attention. Inspired by word embedding (Mikolov et al., 2013), researchers try to solve the link prediction through knowledge graph embedding. Knowledge graph embedding models map entities and rela- tions into low-dimensional vectors (or matrices, tensors), measure the rationality of triples through specific functions between entities and relations, and rank the triples with function scores. Since TransE(Bordes et al., 2013) proposes to use rela- tion vectors to represent the geometric distance between entities, many variants emerge. For exam- ple, TransH(Wang et al., 2014) first explores the different representations of entities under different relations. TransR(Lin et al., 2015) attempts to map entities to the relational space through a particular matrix. TransD(Ji et al., 2015) tries to incorporate the different representations of the entities under the entity and relation into the calculation. These variants attempt to perform complex transforma- tions based on relations or triples to achieve differ- ent representations of entities in different semantic spaces. Recently, scholars are more inclined …

Content truncated to 500 words