Biomedical Entity Linking as Multiple Choice Question Answering

TL;DR

BioELQA introduces a novel approach to biomedical entity linking by framing it as a multiple choice question answering task, improving fine-grained and long-tailed entity recognition.

Contribution

The paper proposes BioELQA, a new model that explicitly compares candidate entities and enhances generalization for long-tailed entities in biomedical linking.

Findings

BioELQA outperforms state-of-the-art baselines on multiple datasets.

Explicit candidate comparison improves fine-grained entity distinction.

Retrieval of similar instances boosts long-tailed entity recognition.

Abstract

Although biomedical entity linking (BioEL) has made significant progress with pre-trained language models, challenges still exist for fine-grained and long-tailed entities. To address these challenges, we present BioELQA, a novel model that treats Biomedical Entity Linking as Multiple Choice Question Answering. BioELQA first obtains candidate entities with a fast retriever, jointly presents the mention and candidate entities to a generator, and then outputs the predicted symbol associated with its chosen entity. This formulation enables explicit comparison of different candidate entities, thus capturing fine-grained interactions between mentions and entities, as well as among entities themselves. To improve generalization for long-tailed entities, we retrieve similar labeled training instances as clues and concatenate the input with retrieved instances for the generator. Extensive experimental results show that BioELQA outperforms state-of-the-art baselines on several datasets.