Learning to Attend On Essential Terms: An Enhanced Retriever-Reader Model for Open-domain Question Answering

By Jianmo Ni, Chenguang Zhu, Weizhu Chen, Julian McAuley. 2018

38 citations recorded by the source corpus.

Open-domain question answering remains a challenging task as it requires models that are capable of understanding questions and answers, collecting useful information, and reasoning over evidence. Previous work typically formulates this task as a reading comprehension or entailment problem given evidence retrieved from search engines. However, existing techniques struggle to retrieve indirectly related evidence when no directly related evidence is provided, especially for complex questions where it is hard to parse precisely what the question asks. In this paper we propose a retriever-reader model that learns to attend on essential terms during the question answering process. We build (1) an essential term selector which first identifies the most important words in a question, then reformulates the query and searches for related evidence; and (2) an enhanced reader that distinguishes between essential terms and distracting words to predict the answer. We evaluate our model on multiple open-domain QA datasets, notably achieving the level of the state-of-the-art on the AI2 Reasoning Challenge (ARC) dataset.

Subfields: Artificial Intelligence, Information Systems, Information Retrieval & Search, Natural Language Processing

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 Information Systems

In Information Retrieval & Search

In Natural Language Processing

Citations in the corpus

Improving Question Answering with External Knowledge (2019, methodology): “Most previous methods (Khashabi et al., 2017; Musa et al., 2018; Ni et al., 2019; Yadav et al., 2019) perform information retrieval on the reference corpus to retrieve relevant sentences to form reference documents.”
Improving Machine Reading Comprehension with General Reading Strategies (2018, methodology): “Approaches marked by Xuse RACE as the source task, except that ET-RR (Ni et al., 2018) uses essential terms (Khashabi et al.”
Answering Any-hop Open-domain Questions with Iterative Document Reranking (2020, background): “Recent works on open-domain QA largely follow the retrieveand-read approach, and havemade prominent improvement on both the retriever module [17, 23, 34] and the reader module [22, 33, 35].”
Think you have Solved Direct-Answer Question Answering? Try ARC-DA, the Direct-Answer AI2 Reasoning Challenge (2021, background): “, (Musa et al., 2019; Boratko et al., 2018; Ni et al., 2019; Xie et al., 2020), and as of January 2021 has 35 entries on its leaderboard2.”
Answering Science Exam Questions Using Query Rewriting with Background Knowledge (2018, methodology): “This module uses a Term Selector which (optionally) incorporates background knowledge in the form of ConceptNet or other embeddings to generate a set of rewritten queries Q = {q1, . . . ,qn}.”
Beyond English-Only Reading Comprehension: Experiments in Zero-shot Multilingual Transfer for Bulgarian (2019, methodology): “, 2018), concatenation and reformulation of passages and questions (Simov et al., 2012; Clark et al., 2016; Ni et al., 2019), as well as re-ranking of documents (Nogueira and Cho, 2019).”
Alignment over Heterogeneous Embeddings for Question Answering (2019, methodology): “Some approaches use query expansion methods in addition to the above methods (Musa et al., 2018; Nogueira and Cho, 2017; Ni et al., 2018).”
Case-Based Abductive Natural Language Inference (2020, result): “At the same time, CB-ANLI obtains competitive results when compared with most of the fine-tuned neural approaches, including ETRR (Ni et al., 2019).”
Large Language Models Can Be Easily Distracted by Irrelevant Context (2023, background): “Jia & Liang (2017) have shown that neural question answering systems are largely affected by adversarial distracting sentences, whereas follow up work (Khashabi et al., 2017; Ni et al., 2019) proposes learning strategies that mitigate the problem.”
Answer Complex Questions: Path Ranker Is All You Need (2021, background): “, 2019 [20]) module in the basic pipeline.”
DDRQA: Dynamic Document Reranking for Open-domain Multi-hop Question Answering (2020, background): “Recent works on open-domain QA largely follow the retrieve-and-read approach, and have made prominent improvement on both the retriever module (Lee et al., 2019; Wang et al., 2018a; Nie et al., 2019) and the reader module (Wang et al., 2018b, 2019; Ni et al., 2019).”