Multi-Passage Machine Reading Comprehension with Cross-Passage Answer Verification

By Yizhong Wang, Kai Liu, Jing Liu, Wei He, Yajuan Lyu, Hua Wu, Sujian Li, Haifeng Wang. 2018

125 citations recorded by the source corpus.

Machine reading comprehension (MRC) on real web data usually requires the machine to answer a question by analyzing multiple passages retrieved by search engine. Compared with MRC on a single passage, multi-passage MRC is more challenging, since we are likely to get multiple confusing answer candidates from different passages. To address this problem, we propose an end-to-end neural model that enables those answer candidates from different passages to verify each other based on their content representations. Specifically, we jointly train three modules that can predict the final answer based on three factors: the answer boundary, the answer content and the cross-passage answer verification. The experimental results show that our method outperforms the baseline by a large margin and achieves the state-of-the-art performance on the English MS-MARCO dataset and the Chinese DuReader dataset, both of which are designed for MRC in real-world settings.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Question Answering & Reading Comprehension

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

In Question Answering & Reading Comprehension

Citations in the corpus

A Multi-Type Multi-Span Network for Reading Comprehension that Requires Discrete Reasoning (2019, neutral): “Morevoer, answer reranking has been well studied in several prior works (Cui et al., 2016; Wang et al., 2018a,b,c; Hu et al., 2019).”
Stochastic Answer Networks for Machine Reading Comprehension (2017, background): “The results in Table 7 show that SAN outperforms V-Net (Wang et al., 2018) and becomes the new state of the art6.”
Errudite: Scalable, Reproducible, and Testable Error Analysis (2019, background): “It has been widely referenced as a strong base-line model (Wang et al., 2018; Clark and Gard-ner, 2017).”
Multi-style Generative Reading Comprehension (2019, methodology): “…find a small number of relevant passages with a TF-IDF based ranker and pass them to a neural reader (Chen et al., 2017; Clark and Gardner, 2018), while more recent pipeline models use a neural re-ranker to more accurately select the relevant passages (Wang et al., 2018; Nishida et al., 2018).”
Subword-augmented Embedding for Cloze Reading Comprehension (2018, background): “Recently, many deep learning models have been proposed for reading comprehension (Sordoni et al., 2016; Trischler et al., 2016; Wang and Jiang, 2016; Munkhdalai and Yu, 2017; Wang et al., 2017a; Dhingra et al., 2017; Zhang et al., 2018b; Wang et al., 2018b).”
Retrieve, Read, Rerank: Towards End-to-End Multi-Document Reading Comprehension (2019, methodology): “Previous approaches such as DS-QA (Lin et al., 2018) and R3 (Wang et al., 2018a) consist of separate retriever and reader models that are jointly trained.”
Quick and (not so) Dirty: Unsupervised Selection of Justification Sentences for Multi-hop Question Answering (2019, methodology): “Unlike other supervised selection methods (Dehghani et al., 2019; Bao et al., 2016; Lin et al., 2018; Wang et al., 2018b,a; Tran and Niedereée, 2018; Trivedi et al., 2019), our approach does not require any training data for justification selection.”
BiPaR: A Bilingual Parallel Dataset for Multilingual and Cross-lingual Reading Comprehension on Novels (2019, background): “It has been regarded as a crucial technology for many applications such as question answering, dialogue systems ( Nguyen et al., 2016; Chen et al 2017; Liu et al., 2018; Wang et al., 2018) and so on.”
Effective Subword Segmentation for Text Comprehension (2018, background): “With the release of the large-scale span-based datasets [21], [35], [39], [40], [48], which constrain answers to all possible”
RECONSIDER: Re-Ranking using Span-Focused Cross-Attention for Open Domain Question Answering (2020, methodology): “This contrasts with multi-task training approaches (Wang et al., 2018b),”
Human Behavior Inspired Machine Reading Comprehension (2019, unknown): “Because most of the wisdom teeth grow out between the ages of sixteen and twenty-four, it appears after the human IQ is mature, so it is also known as the wisdom tooth Recently, several neural MRC models [8, 30, 31, 35, 37, 38] have been proposed with the inspiration from human reading behavior.”
I Know What You Want: Semantic Learning for Text Comprehension (2018, background): “Wang et al. 2018; Rajpurkar, Jia, and Liang 2018), which constrain answers to all possible text spans within the reference document, researchers are investigating the models with more logical reasoning and content understanding (Wang et al. 2018; Wang, Yan, and Wu 2018).”