ImCoref-CeS: An Improved Lightweight Pipeline for Coreference Resolution with LLM-based Checker-Splitter Refinement

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Luo, Kangyang, Bai, Yuzhuo, Si, Shuzheng, Gao, Cheng, Wang, Zhitong, Shen, Yingli, Li, Wenhao, Liu, Zhu, Han, Yufeng, Wu, Jiayi, Kong, Cunliang, Sun, Maosong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911653780520960
author Luo, Kangyang
Bai, Yuzhuo
Si, Shuzheng
Gao, Cheng
Wang, Zhitong
Shen, Yingli
Li, Wenhao
Liu, Zhu
Han, Yufeng
Wu, Jiayi
Kong, Cunliang
Sun, Maosong
author_facet Luo, Kangyang
Bai, Yuzhuo
Si, Shuzheng
Gao, Cheng
Wang, Zhitong
Shen, Yingli
Li, Wenhao
Liu, Zhu
Han, Yufeng
Wu, Jiayi
Kong, Cunliang
Sun, Maosong
contents Coreference Resolution (CR) is a critical task in Natural Language Processing (NLP). Current research faces a key dilemma: whether to further explore the potential of supervised neural methods based on small language models, whose detect-then-cluster pipeline still delivers top performance, or embrace the powerful capabilities of Large Language Models (LLMs). However, effectively combining their strengths remains underexplored. To this end, we propose \textbf{ImCoref-CeS}, a novel framework that integrates an enhanced supervised model with LLM-based reasoning. First, we present an improved CR method (\textbf{ImCoref}) to push the performance boundaries of the supervised neural method by introducing a lightweight bridging module to enhance long-text encoding capability, devising a biaffine scorer to comprehensively capture positional information, and invoking a hybrid mention regularization to improve training efficiency. Importantly, we employ an LLM acting as a multi-role Checker-Splitter agent to validate candidate mentions (filtering out invalid ones) and coreference results (splitting erroneous clusters) predicted by ImCoref. Extensive experiments demonstrate the effectiveness of ImCoref-CeS, which achieves superior performance compared to existing state-of-the-art (SOTA) methods.
format Preprint
id arxiv_https___arxiv_org_abs_2510_10241
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ImCoref-CeS: An Improved Lightweight Pipeline for Coreference Resolution with LLM-based Checker-Splitter Refinement
Luo, Kangyang
Bai, Yuzhuo
Si, Shuzheng
Gao, Cheng
Wang, Zhitong
Shen, Yingli
Li, Wenhao
Liu, Zhu
Han, Yufeng
Wu, Jiayi
Kong, Cunliang
Sun, Maosong
Computation and Language
Information Retrieval
Coreference Resolution (CR) is a critical task in Natural Language Processing (NLP). Current research faces a key dilemma: whether to further explore the potential of supervised neural methods based on small language models, whose detect-then-cluster pipeline still delivers top performance, or embrace the powerful capabilities of Large Language Models (LLMs). However, effectively combining their strengths remains underexplored. To this end, we propose \textbf{ImCoref-CeS}, a novel framework that integrates an enhanced supervised model with LLM-based reasoning. First, we present an improved CR method (\textbf{ImCoref}) to push the performance boundaries of the supervised neural method by introducing a lightweight bridging module to enhance long-text encoding capability, devising a biaffine scorer to comprehensively capture positional information, and invoking a hybrid mention regularization to improve training efficiency. Importantly, we employ an LLM acting as a multi-role Checker-Splitter agent to validate candidate mentions (filtering out invalid ones) and coreference results (splitting erroneous clusters) predicted by ImCoref. Extensive experiments demonstrate the effectiveness of ImCoref-CeS, which achieves superior performance compared to existing state-of-the-art (SOTA) methods.
title ImCoref-CeS: An Improved Lightweight Pipeline for Coreference Resolution with LLM-based Checker-Splitter Refinement
topic Computation and Language
Information Retrieval
url https://arxiv.org/abs/2510.10241