arXiv cs.CL· Zhuoqi He, Peijie Huang, Yuhong Xu, Youming Peng, Mingzhi Xu, Xinyang Lin·· 13 小时前AI 评分22
MLEGT：用多级编码器与全局主题实现任务型对话摘要
Multi-level Encoder with Global Topic for Task-oriented Dialogue Summarization
AI 导读
研究者提出任务型对话摘要模型 MLEGT，通过多级编码器捕捉对话的时序与空间依赖，并在图构建过程中获得简洁有效的全局主题，从而突破长距离限制、发现对话内在结构。摘要器还引入全局主题引导的指针机制，用于捕捉此前被忽略但重要的信息以提升摘要准确率。在真实大规模中文审讯数据集和医疗报告服务数据集两个公开数据集上，该方法优于多个强基线。
正文 · 原文
Abstract:Task-oriented dialogue summarization aims to automatically extract key information to generate domain summaries to improve service efficiency and quality. Task-oriented dialogue is inherently logical and surrounds specific topic. How to effectively capture the dialogue topic and the most salient information becomes one of the major challenges of this task. In this paper, we propose a task-oriented dialogue summarization model utilizing a multi-level encoder with global topic (MLEGT). This model discovers temporal and spatial dependencies through a multi-level encoder, while simultaneously obtaining a simple yet effective global topic in the process of graph construction. It not only captures the crucial details but also breaks the long-range limit and discover the intrinsic structure across the dialogue. Then, a global topic guided pointer mechanism is added into the summrizer to capture ignored but important information, which is helpful to improve the accuracy of the summary. A comprehensive study of two public datasets, including a real-world large-scale Chinese police interrogation dataset and a medical reporting service dataset proves the superiority of our method on several strong baselines.
| Comments: | 5 pages, 3 figures. Published in ICASSP 2025 |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.29097 [cs.CL] |
| (or arXiv:2610.29097v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.29097 arXiv-issued DOI via DataCite (pending registration) |
|
| Journal reference: | Proc. IEEE ICASSP 2025, pp. 1-5 (2025) |
| Related DOI: | https://doi.org/10.1109/ICASSP49660.2025.10890255
DOI(s) linking to related resources |
Submission history
From: Zhuoqi He [view email]
[v1]
Wed, 21 Oct 2026 02:41:38 UTC (1,134 KB)
来源：arXiv cs.CL · arxiv.org