Abstract:Multi-passage machine reading comprehension (MRC) aims to answer a question by multiple passages. Existing multi-passage MRC approaches have shown that employing passages with and without golden answers (i.e. labeled and unlabeled passages) for model training can improve prediction accuracy. However, when using the unlabeled passages, they either incur the wrong labeling problem or treat the labeled and unlabeled passages equally. In addition, they ignore the original passage information to verify the correctness of the answer. In this paper, we present MLDV-MRC, a novel approach for multi-passage MRC via M ulti-task L earning and D ual V erification. MLDV-MRC adopts the extract-then-select framework, where an extractor is first used to predict answer candidates, then a selector is used to choose the final answer. For the extractor, we adopt multi-task learning with generative adversarial training to train it by using both labeled and unlabeled passages. To train the extractor by backpropagation, we propose a hybrid method which combines boundary-based and content-based extracting methods to produce the answer candidate set and its representation. For the selector, we propose to leverage both the information from answer candidates and original passages to verify the final answer. In particular, we propose a global-local memory-augmented neural network to build the representations of original passages, which fuses the passage-level information and word-level information. The experimental results on three open-domain QA datasets confirm the effectiveness of our approach.

A Multi-answer Multi-task Framework for Real-world Machine Reading Comprehension.

Hierarchical Answer Selection Framework For Multi-Passage Machine Reading Comprehension

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

Multi-Task Deep Neural Networks for Multi-Document Reading Comprehension

How Many Answers Should I Give? an Empirical Study of Multi-Answer Reading Comprehension.

Multi-Matching Network for Multiple Choice Reading Comprehension

Multi-Passage Machine Reading Comprehension Through Multi-Task Learning and Dual Verification

Feeding What You Need by Understanding What You Learned

Multi-choice Dialogue-Based Reading Comprehension with Knowledge and Key Turns.

Multiple Perspective Answer Reranking for Multi-passage Reading Comprehension.

Machine Reading Comprehension for Continuous Questions Via Multi-Task Learning

MMM: Multi-stage Multi-task Learning for Multi-choice Reading Comprehension

Enhancing Machine Comprehension Using Multi-Knowledge Bases and Offline Answer Span Improving System

A Multi-Task Learning Machine Reading Comprehension Model for Noisy Document (student Abstract)

Coarse-to-Fine Document Ranking for Multi-Document Reading Comprehension with Answer-Completion.

A Sentence Quality Evaluation Framework for Machine Reading Comprehension Incorporating Pre-trained Language Model.

Integrate Candidate Answer Extraction with Re-Ranking for Chinese Machine Reading Comprehension

Selecting Paragraphs to Answer Questions for Multi-passage Machine Reading Comprehension

MA-MRC: A Multi-answer Machine Reading Comprehension Dataset

Multi-Strategy to Improve Reading Comprehension

Read, Attend, and Exclude: Multi-Choice Reading Comprehension by Mimicking Human Reasoning Process