Knowledge-Based Question Answering as Machine Translation

By Junwei Bao, Nan Duan, M. Zhou, Tuo Zhao. 2014

136 citations recorded by the source corpus.

A typical knowledge-based question answering (KB-QA) system faces two challenges: one is to transform natural language questions into their meaning representations (MRs); the other is to retrieve answers from knowledge bases (KBs) using generated MRs. Unlike previous methods which treat them in a cascaded manner, we present a translation-based approach to solve these two tasks in one unified framework. We translate questions to answers based on CYK parsing. Answers as translations of the span covered by each CYK cell are obtained by a question translation method, which first generates formal triple queries as MRs for the span based on question patterns and relation expressions, and then retrieves answers from a given KB based on triple queries generated. A linear model is defined over derivations, and minimum error rate training is used to tune feature weights based on a set of question-answer pairs. Compared to a KB-QA system using a state-of-the-art semantic parser, our method achieves better results.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Semantic Parsing via Staged Query Graph Generation: Question Answering with Knowledge Base (2015, methodology): “Inspired by (Yao and Van Durme, 2014; Bao et al., 2014), we propose a semantic parsing framework that leverages the knowledge base more tightly when forming the parse for an input question.”
Question Answering over Freebase with Multi-Column Convolutional Neural Networks (2015, background): “P@1 (Berant et al., 2013) 31.4 - (Berant and Liang, 2014) 39.9 - (Bao et al., 2014) 37.5 - (Yao and Van Durme, 2014) 33.0 - (Bordes et al., 2014a) 39.2 40.4 (Bordes et al., 2014b) 29.7 31.3 MCCNN (our) 40.8 45.1 Table 1: Evaluation results on the test split of W E -”
Transforming Dependency Structures to Logical Forms for Semantic Parsing (2016, unknown): “9 Bao et al. (2014) – 37.”
Neural Approaches to Conversational AI (2018, unknown): “estion can then be obtained by ﬁnding a set of paths in the KB that match the query and retrieving the end nodes of these paths (Richardson et al., 1998; Berant et al., 2013; Yao and Van Durme, 2014; Bao et al., 2014; Yih et al., 2015b). We take the example used in Yih et al. (2015a) to illustrate t”
Using Local Knowledge Graph Construction to Scale Seq2Seq Models to Multi-Document Inputs (2019, background): “Knowledge bases have been leveraged to improve performance on various tasks, from coreference resolution (Ng and Cardie, 2002) and question answering (Zheng, 2003; Bao et al., 2014; Cui et al., 2017; Sun et al., 2018) to signal processing (Bückner et al.”
Learning to Diagnose: Assimilating Clinical Narratives using Deep Reinforcement Learning (2017, background): “Addressing inference tasks generally requires signiﬁcant contributions from domain experts and access to a variety of resources (Ferrucci et al., 2013; Lally et al., 2014) such as structured knowledge bases (KBs) (Yao and Van Durme, 2014; Bao et al., 2014; Dong et al., 2015).”
Statistical learning for semantic parsing: A survey (2019, background): “[61, 62], STAGG[60] proposed a semantic parsing that leverages the knowledge base tightly forming the parse for an input question.”
Scene Graph Generation With External Knowledge and Image Reconstruction (2019, background): “There has been growing interest in improving data-driven models with external Knowledge Bases (KBs) in natural language processing [17, 4] and computer vision communities [24, 1, 6].”
Question Answering on Freebase via Relation Extraction and Textual Evidence (2016, methodology): “Inspired by Bao et al. (2014), we design a dependency treebased method to handle such multi-relational questions.”
Medical Question Answering for Clinical Decision Support (2016, methodology): “With the recent growth of KBs such as DBPedia [3] and Freebase [7], new promising methods for Q/A from KBs have emerged [9, 34, 5].”