🧠 LLM-based Persuasion Simulator
Official model and repository for the following papers:
Communication Makes Perfect: Persuasion Dataset Construction via Multi-LLM Communication
NAACL 2025 (Main Conference)
https://aclanthology.org/2025.naacl-long.203/
Communication is All You Need: Persuasion Dataset Construction via Multi-LLM Communication
arXiv preprint, February 2025
arXiv:2502.08896
Both versions describe the same multi-LLM communication framework for generating persuasive dialogue datasets,
with the NAACL 2025 paper being the peer-reviewed and finalized version.
📘 Overview
This repository provides the LLM-based persuasion simulation framework introduced in our paper.
The framework enables automated generation of high-quality persuasive dialogues through multi-agent communication among large language models (LLMs),
covering diverse linguistic and strategic dimensions of persuasion.
Key features:
- Multi-LLM communication to enhance dialogue fluency, coherence, and diversity
- Fine-grained control over persuasion strategies
- Modular design for flexible extension and reproducibility
- Applied to construct the PersuaSim and PersuaSim-Culture datasets
📂 Code Repository
💻 Full implementation and scripts are available at:
👉 https://github.com/HF-heaven/LLM-based_persuasion_simulator
🧩 Citation
If you find this work useful, please cite:
@inproceedings{ma2025communication,
title={Communication Makes Perfect: Persuasion Dataset Construction via Multi-LLM Communication},
author={Ma, Weicheng and Zhang, Hefan and Yang, Ivory and Ji, Shiyu and Chen, Joice and Hashemi, Farnoosh and Mohole, Shubham and Gearey, Ethan and Macy, Michael and Hassanpour, Saeed and others},
booktitle={Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)},
pages={4017--4045},
year={2025}
}