Proxy-based Zero-Shot Entity Linking by Effective Candidate Retrieval

TL;DR

This paper introduces a proxy-based metric learning approach with adversarial regularization for efficient candidate retrieval in biomedical entity linking, enabling high recall and zero-shot discovery without costly ranking steps.

Contribution

It proposes a novel proxy-based metric learning method combined with adversarial regularization to improve candidate retrieval efficiency in biomedical entity linking.

Findings

Achieves competitive recall@1 performance

Eliminates need for expensive candidate ranking step

Enables zero-shot out-of-knowledge-base entity discovery

Abstract

A recent advancement in the domain of biomedical Entity Linking is the development of powerful two-stage algorithms, an initial candidate retrieval stage that generates a shortlist of entities for each mention, followed by a candidate ranking stage. However, the effectiveness of both stages are inextricably dependent on computationally expensive components. Specifically, in candidate retrieval via dense representation retrieval it is important to have hard negative samples, which require repeated forward passes and nearest neighbour searches across the entire entity label set throughout training. In this work, we show that pairing a proxy-based metric learning loss with an adversarial regularizer provides an efficient alternative to hard negative sampling in the candidate retrieval stage. In particular, we show competitive performance on the recall@1 metric, thereby providing the option to leave out the expensive candidate ranking step. Finally, we demonstrate how the model can be used in a zero-shot setting to discover out of knowledge base biomedical entities.