Direct Fact Retrieval from Knowledge Graphs without Entity Linking

Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 10038–10055 July 9-14, 2023 ©2023 Association for Computational Linguistics Direct Fact Retrieval from Knowledge Graphs without Entity Linking Jinheon Baek1∗ Alham Fikri Aji2 Jens Lehmann3 Sung Ju Hwang1 KAIST1 MBZUAI2 Amazon3 {jinheon.baek, sjhwang82}@kaist.ac.kr alham.fikri@mbzuai.ac.ae jlehmnn@amazon.com Abstract There has been a surge of interest in utilizing Knowledge Graphs (KGs) for various natural language processing/understanding tasks. The conventional mechanism to retrieve facts in KGs usually involves three steps: entity span detection, entity disambiguation, and relation classification. However, this approach requires additional labels for training each of the three subcomponents in addition to pairs of input texts and facts, and also may accumulate errors propagated from failures in previous steps. To tackle these limitations, we propose a simple knowledge retrieval framework, which directly retrieves facts from the KGs given the input text based on their representational similarities, which we refer to as Direct Fact Retrieval (Di- FaR). Specifically, we first embed all facts in KGs onto a dense embedding space by using a language model trained by only pairs of input texts and facts, and then provide the nearest facts in response to the input text. Since the fact, consisting of only two entities and one re- lation, has little context to encode, we propose to further refine ranks of top-k retrieved facts with a reranker that contextualizes the input text and the fact jointly. We validate our Di- FaR framework on multiple fact retrieval tasks, showing that it significantly outperforms rele- vant baselines that use the three-step approach. 1 Introduction Knowledge graphs (KGs) (Bollacker et al., 2008; Vrandecic and Krötzsch, 2014; Lehmann et al., 2015), which consist of a set of facts represented in the form of a (head entity, relation, tail entity) triplet, can store a large amount of world knowl- edge. In natural language applications, language models (LMs) (Devlin et al., 2019; Brown et al., 2020) are commonly used; however, their knowl- edge internalized in parameters is often incomplete, inaccurate, and outdated. Therefore, several recent ∗ Work done while interning at Amazon. Corresponding author: Jinheon Baek (jinheon.baek@kaist.ac.kr) Q: Where was Michael F. Phelps born? (a) Conventional Fact Retrieval (b) Our Direct Fact Retrieval Michael F. Phelps (1) Entity Mention Detection Q39562 (Michael Phelps) (2) Entity Disambiguation IDs in the WIKIDATA [Michael Phelps, Place of birth, Baltimore] Michael Phelps (3) Relation Classification Baltimore Nicole Johnson Spouse Swimming Sport Q: Where was Michael F. Phelps born? [Michael Phelps, Place of birth, Baltimore] Michael Phelps Baltimore Place_of_birth Michael Phelps Baltimore Sport Facts in a Knowledge Graph Nicole Johnson Model Occupation … A representation space by one language model Direct Retrieval without Entity Linking Figure 1: (a) A conventional fact retrieval from KGs involves three sequential steps: 1) entity mention detection to identify entities in queries; 2) entity disambiguation to match entities in input texts to KGs; 3) relation classification to select relevant relations. (b) Our fact retrieval directly retrieves relevant facts with their representational similarities to …

Content truncated to 500 words