Tracing Origins: Coreference-aware Machine Reading Comprehension

TL;DR

This paper enhances machine reading comprehension by explicitly incorporating coreference information into pre-trained language models, improving performance on coreference-intensive tasks like question answering.

Contribution

It introduces two fine-tuning strategies that explicitly leverage coreference information, outperforming methods that incorporate such info during pre-training.

Findings

Explicit coreference incorporation improves model performance.

Fine-tuning with coreference info outperforms pre-training methods.

Graph convolutional networks effectively model coreference relations.

Abstract

Machine reading comprehension is a heavily-studied research and test field for evaluating new pre-trained language models (PrLMs) and fine-tuning strategies, and recent studies have enriched the pre-trained language models with syntactic, semantic and other linguistic information to improve the performance of the models. In this paper, we imitate the human reading process in connecting the anaphoric expressions and explicitly leverage the coreference information of the entities to enhance the word embeddings from the pre-trained language model, in order to highlight the coreference mentions of the entities that must be identified for coreference-intensive question answering in QUOREF, a relatively new dataset that is specifically designed to evaluate the coreference-related performance of a model. We use two strategies to fine-tune a pre-trained language model, namely, placing an additional encoder layer after a pre-trained language model to focus on the coreference mentions or constructing a relational graph convolutional network to model the coreference relations. We demonstrate that the explicit incorporation of coreference information in the fine-tuning stage performs better than the incorporation of the coreference information in pre-training a language model.