Answer Span Correction in Machine Reading Comprehension

TL;DR

This paper introduces a post-processing correction method for machine reading comprehension systems to improve answer accuracy, addressing partial answer errors and demonstrating significant performance gains across multiple languages.

Contribution

It presents a novel answer span correction technique that enhances existing MRC models by reducing partial answer errors through post-processing.

Findings

Significant performance improvements over state-of-the-art MRC systems.

Effective in both monolingual and multilingual settings.

Addresses partial answer errors in MRC outputs.

Abstract

Answer validation in machine reading comprehension (MRC) consists of verifying an extracted answer against an input context and question pair. Previous work has looked at re-assessing the "answerability" of the question given the extracted answer. Here we address a different problem: the tendency of existing MRC systems to produce partially correct answers when presented with answerable questions. We explore the nature of such errors and propose a post-processing correction method that yields statistically significant performance improvements over state-of-the-art MRC systems in both monolingual and multilingual evaluation.