Skip to content

ElPlaguister/Def-DTS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Def-DTS

Official code for ACL 2025 paper: Def-DTS: Deductive Reasoning for Open-domain Dialogue Topic Segmentation.

Overview

Image

This is the overview of Def-DTS framework.

  • We introduce LLM reasoning techniques to DTS for the first time, consolidating insights from previous methodologies into a coherent and deductive prompt design.

  • We reformulate DTS as an utterance-level intent classification task by implementing intent classification as a core component of a multi-step reasoning process, enabling flexible and task-agnostic prompting.

  • Our method empirically demonstrates superior performance across nearly all comparative baselines, underscoring the efficacy of prompt engineering in DTS.

  • Through an in-depth analysis of our approach's reasoning results, we shed light on the challenges LLM reasoning faces in DTS and discuss the possibility of using LLM as a DTS auto-labeler.

Please refer to the paper for details.

Citation

@misc{lee2025defdts,
  title={Def-DTS: Deductive Reasoning for Open-domain Dialogue Topic Segmentation},
  author={Seungmin Lee and Yongsang Yoo and Minhwa Jung and Min Song},
  year={2025},
  eprint={2505.21033},
  archivePrefix={arXiv},
  primaryClass={cs.CL}
}

About

Findings of the ACL 2025

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published