Fast and Effective Biomedical Entity Linking Using a Dual Encoder
This addresses efficiency bottlenecks in biomedical entity linking for researchers and practitioners, offering a faster alternative with competitive performance, though it is incremental as it builds on existing BERT-based paradigms.
The paper tackles the slow training and inference speed of BERT-based biomedical entity linking models by proposing a dual encoder that processes multiple mentions in one shot, achieving competitive accuracy while being multiple times faster. It also extends this to an end-to-end model that outperforms recent methods in span detection and disambiguation.
Biomedical entity linking is the task of identifying mentions of biomedical concepts in text documents and mapping them to canonical entities in a target thesaurus. Recent advancements in entity linking using BERT-based models follow a retrieve and rerank paradigm, where the candidate entities are first selected using a retriever model, and then the retrieved candidates are ranked by a reranker model. While this paradigm produces state-of-the-art results, they are slow both at training and test time as they can process only one mention at a time. To mitigate these issues, we propose a BERT-based dual encoder model that resolves multiple mentions in a document in one shot. We show that our proposed model is multiple times faster than existing BERT-based models while being competitive in accuracy for biomedical entity linking. Additionally, we modify our dual encoder model for end-to-end biomedical entity linking that performs both mention span detection and entity disambiguation and out-performs two recently proposed models.