KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain Question Answering

Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 4961 - 4974 May 22-27, 2022 c⃝2022 Association for Computational Linguistics KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain Question Answering Donghan Yu1∗, Chenguang Zhu2, Yuwei Fang2, Wenhao Yu3∗, Shuohang Wang2, Yichong Xu2, Xiang Ren4, Yiming Yang1, Michael Zeng2 1Carnegie Mellon University 2Microsoft Cognitive Services Research Group 3University of Notre Dame 4University of Southern California 1dyu2@cs.cmu.edu, 2chezhu@microsoft.com Abstract Current Open-Domain Question Answering (ODQA) models typically include a retriev- ing module and a reading module, where the retriever selects potentially relevant passages from open-source documents for a given ques- tion, and the reader produces an answer based on the retrieved passages. The recently pro- posed Fusion-in-Decoder (FiD) framework is a representative example, which is built on top of a dense passage retriever and a gener- ative reader, achieving the state-of-the-art per- formance. In this paper we further improve the FiD approach by introducing a knowledge- enhanced version, namely KG-FiD. Our new model uses a knowledge graph to establish the structural relationship among the retrieved pas- sages, and a graph neural network (GNN) to re-rank the passages and select only a top few for further processing. Our experiments on common ODQA benchmark datasets (Natural Questions and TriviaQA) demonstrate that KG- FiD can achieve comparable or better perfor- mance in answer prediction than FiD, with less than 40% of the computation cost. 1 Introduction Open-Domain Question Answering (ODQA) is the task of answering natural language questions in open domains. A successful ODQA model relies on effective acquisition of world knowledge. A popular line of work treats a large collection of open-domain documents (such as Wikipedia arti- cles) as the knowledge source, and design a ODQA system that consists of a retrieving module and a reading module. The retriever pulls out a small set of potentially relevant passages from the open- source documents for a given question, and the reader produces an answer based on the retrieved passages (Karpukhin et al., 2020; Guu et al., 2020; Izacard and Grave, 2020). An earlier example of this kind is DrQA (Chen et al., 2017), which used ∗Work done during internship at Microsoft. an traditional search engine based on the bag of words (BoW) document representation with TF- IDF term weighting, and a neural reader for extract- ing candidate answers for each query based on the dense embedding of the retrieved passages. With the successful development of Pre-trained Lan- guage Models (PLMs) in neural network research, dense embedding based passage retrieval (DPR) models (Karpukhin et al., 2020; Qu et al., 2021) have shown superior performance over BoW/TF- IDF based retrieval models due to utilization of contextualized word embedding in DPR, and gen- erative QA readers (Lewis et al., 2020; Roberts et al., 2020) usually outperform extraction based readers (Devlin et al., 2019; Guu et al., 2020) due to the capability of the former in capturing lexical variants with a richer flexibility. The recently proposed Fusion-in-Decoder (FiD) model (Izacard and Grave, …

Content truncated to 500 words