Empirical Evaluation of Pretraining Strategies for Supervised Entity Linking
doi:10.24432/C59G6S
TL;DR
We achieve state of the art on CoNLL and TAC-KBP 2010 with a four layer transformer In this work, we present an entity linking model which combines a Transformer architecture with large scale pretraining from Wikipedia links. Our model achieves the state-of-the-art on two commonly used entity linking datasets: 96.7% on CoNLL and 94.9% on TAC-KBP. We present detailed analyses to understand what design choices are important for entity linking, including choices of negative entity candidates, Transformer architecture, and input perturbations. Lastly, we present promising results on more challenging settings such as end-to-end entity linking and entity linking without in-domain training data
Citation
@inproceedings{
f{\'e}vry2020empirical,
title={Empirical Evaluation of Pretraining Strategies for Supervised Entity Linking},
author={Thibault F{\'e}vry and Nicholas FitzGerald and Livio Baldini Soares and Tom Kwiatkowski},
booktitle={Automated Knowledge Base Construction},
year={2020},
url={https://openreview.net/forum?id=iHXV8UGYyL},
doi={10.24432/C59G6S}
}
