Improving Biomedical Entity Linking with Retrieval-enhanced Learning

Abstract

Biomedical entity linking (BioEL) has achieved remarkable progress with the help of pre-trained language models. However, existing BioEL methods usually struggle to handle rare and difficult entities due to long-tailed distribution. To address this limitation, we introduce a new scheme $k$NN-BioEL, which provides a BioEL model with the ability to reference similar instances from the entire training corpus as clues for prediction, thus improving the generalization capabilities. Moreover, we design a contrastive learning objective with dynamic hard negative sampling (DHNS) that improves the quality of the retrieved neighbors during inference. Extensive experimental results show that $k$NN-BioEL outperforms state-of-the-art baselines on several datasets.