Joint Information Extraction Across Classical and Modern Chinese with Tea-MOELoRA

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tang, Xuemei, Yan, Chengxi, Gu, Jinghang, Huang, Chu-Ren
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918138025607168
author Tang, Xuemei
Yan, Chengxi
Gu, Jinghang
Huang, Chu-Ren
author_facet Tang, Xuemei
Yan, Chengxi
Gu, Jinghang
Huang, Chu-Ren
contents Chinese information extraction (IE) involves multiple tasks across diverse temporal domains, including Classical and Modern documents. Fine-tuning a single model on heterogeneous tasks and across different eras may lead to interference and reduced performance. Therefore, in this paper, we propose Tea-MOELoRA, a parameter-efficient multi-task framework that combines LoRA with a Mixture-of-Experts (MoE) design. Multiple low-rank LoRA experts specialize in different IE tasks and eras, while a task-era-aware router mechanism dynamically allocates expert contributions. Experiments show that Tea-MOELoRA outperforms both single-task and joint LoRA baselines, demonstrating its ability to leverage task and temporal knowledge effectively.
format Preprint
id arxiv_https___arxiv_org_abs_2509_01158
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Joint Information Extraction Across Classical and Modern Chinese with Tea-MOELoRA
Tang, Xuemei
Yan, Chengxi
Gu, Jinghang
Huang, Chu-Ren
Computation and Language
Chinese information extraction (IE) involves multiple tasks across diverse temporal domains, including Classical and Modern documents. Fine-tuning a single model on heterogeneous tasks and across different eras may lead to interference and reduced performance. Therefore, in this paper, we propose Tea-MOELoRA, a parameter-efficient multi-task framework that combines LoRA with a Mixture-of-Experts (MoE) design. Multiple low-rank LoRA experts specialize in different IE tasks and eras, while a task-era-aware router mechanism dynamically allocates expert contributions. Experiments show that Tea-MOELoRA outperforms both single-task and joint LoRA baselines, demonstrating its ability to leverage task and temporal knowledge effectively.
title Joint Information Extraction Across Classical and Modern Chinese with Tea-MOELoRA
topic Computation and Language
url https://arxiv.org/abs/2509.01158