Mulco: Recognizing Chinese Nested Named Entities Through Multiple Scopes

By Jiuding Yang et al

Read the original document by opening this link in a new tab.

Table of Contents

Abstract
Introduction
Related Works
Dataset
Methodology

Summary

Nested Named Entity Recognition (NNER) has been a long-term challenge to researchers as an important sub-area of Named Entity Recognition. This paper focuses on Chinese Nested Named Entity Recognition (CNNER) and introduces Mulco, a novel method that can recognize named entities in nested structures through multiple scopes. The proposed approach uses scope-based sequence labeling to identify named entities accurately. The method outperforms several baseline methods in experiments and is specially designed for CNNER. The ChiNesE dataset is constructed to facilitate research on CNNER, offering a larger and more diverse dataset compared to ACE 2005. Mulco utilizes scopes inspired by modern Computer Vision methods to locate entities in sentences and employs a sequence labeling scheme for recognition.