Bidirectional End-to-End Learning of Retriever-Reader Paradigm for Entity Linking

TL;DR

This paper introduces BEER$^2$, a bidirectional end-to-end training framework that enhances entity linking by enabling mutual learning between retriever and reader components, leading to improved performance.

Contribution

The paper proposes a novel bidirectional end-to-end training method for retriever and reader in entity linking, allowing mutual learning and outperforming existing pipeline approaches.

Findings

BEER$^2$ improves entity linking accuracy across multiple benchmarks.

Bidirectional training enhances the interaction between retriever and reader.

Experimental results show significant performance gains.

Abstract

Entity Linking (EL) is a fundamental task for Information Extraction and Knowledge Graphs. The general form of EL (i.e., end-to-end EL) aims to first find mentions in the given input document and then link the mentions to corresponding entities in a specific knowledge base. Recently, the paradigm of retriever-reader promotes the progress of end-to-end EL, benefiting from the advantages of dense entity retrieval and machine reading comprehension. However, the existing study only trains the retriever and the reader separately in a pipeline manner, which ignores the benefit that the interaction between the retriever and the reader can bring to the task. To advance the retriever-reader paradigm to perform more perfectly on end-to-end EL, we propose BEER$^2$, a Bidirectional End-to-End training framework for Retriever and Reader. Through our designed bidirectional end-to-end training, BEER$^2$ guides the retriever and the reader to learn from each other, make progress together, and ultimately improve EL performance. Extensive experiments on benchmarks of multiple domains demonstrate the effectiveness of our proposed BEER$^2$.