Reference Knowledgeable Network for Machine Reading Comprehension

TL;DR

RekNet is a novel model for multi-choice machine reading comprehension that enhances understanding by explicitly referencing external knowledge sources and simulating human reading strategies, leading to improved performance.

Contribution

The paper introduces RekNet, a new model that refines critical information and explicitly incorporates external knowledge, addressing limitations of previous models that lacked external knowledge integration.

Findings

Achieved significant performance improvements on RACE, DREAM, and Cosmos QA benchmarks.

Demonstrated the effectiveness of explicit external knowledge referencing in MRC.

Statistically significant results over strong baseline models.

Abstract

Multi-choice Machine Reading Comprehension (MRC) as a challenge requires models to select the most appropriate answer from a set of candidates with a given passage and question. Most of the existing researches focus on the modeling of specific tasks or complex networks, without explicitly referring to relevant and credible external knowledge sources, which are supposed to greatly make up for the deficiency of the given passage. Thus we propose a novel reference-based knowledge enhancement model called Reference Knowledgeable Network (RekNet), which simulates human reading strategies to refine critical information from the passage and quote explicit knowledge in necessity. In detail, RekNet refines finegrained critical information and defines it as Reference Span, then quotes explicit knowledge quadruples by the co-occurrence information of Reference Span and candidates. The proposed RekNet is evaluated on three multi-choice MRC benchmarks: RACE, DREAM and Cosmos QA, obtaining consistent and remarkable performance improvement with observable statistical significance level over strong baselines. Our code is available at https://github.com/Yilin1111/RekNet.