An Empirical Study of Representation, Training and Decoding for Span-based Named Entity Recognition

Abstract

Named Entity Recognition (NER) is an important task in Natural Language Processing with application in many domains. While the dominant paradigm of NER is sequence labelling, span-based approaches have become very popular in recent times, but are less well understood. In this work, we study different aspects of span-based NER, namely the span representation, learning strategy, and decoding algorithms to avoid span overlap. We also propose an exact algorithm that efficiently finds the set of non-overlapping spans that maximize a global score, given a list of candidate spans. We perform our study on three benchmarks NER datasets from different domains. The code and supporting files for the experiments will be made publicly available.

Tasks

Reproductions