Fusing Heterogeneous Factors with Triaffine Mechanism for Nested Named Entity Recognition

TL;DR

This paper introduces a novel triaffine mechanism to effectively fuse heterogeneous factors for nested named entity recognition, significantly improving span classification performance and achieving state-of-the-art results on multiple datasets.

Contribution

It proposes a new triaffine mechanism for integrating diverse span-related factors, enhancing nested NER accuracy beyond existing span-based methods.

Findings

Outperforms previous span-based methods on GENIA and KBP2017 datasets.

Achieves state-of-the-art F1 scores for nested NER.

Shows comparable results on ACE2004 and ACE2005 datasets.

Abstract

Nested entities are observed in many domains due to their compositionality, which cannot be easily recognized by the widely-used sequence labeling framework. A natural solution is to treat the task as a span classification problem. To learn better span representation and increase classification performance, it is crucial to effectively integrate heterogeneous factors including inside tokens, boundaries, labels, and related spans which could be contributing to nested entities recognition. To fuse these heterogeneous factors, we propose a novel triaffine mechanism including triaffine attention and scoring. Triaffine attention uses boundaries and labels as queries and uses inside tokens and related spans as keys and values for span representations. Triaffine scoring interacts with boundaries and span representations for classification. Experiments show that our proposed method outperforms previous span-based methods, achieves the state-of-the-art $F_1$ scores on nested NER datasets GENIA and KBP2017, and shows comparable results on ACE2004 and ACE2005.