MARO is an end-to-end multi-agent framework that empowers Large Language Models (LLMs) to detect cross-domain misinformation through automated analysis and optimized decision-making.
Overview of the MARO frameworkPlease configure your environment and add your own API key and prompts in the code before running the Quick Start instructions below.
python MARO\report.pypython MARO\INS.pypython MARO\test.pyPlease cite the following if you find MARO helpful:
@article{DBLP:journals/corr/abs-2503-23329,
publtype={informal},
author={Hui Li and Ante Wang and Kunquan Li and Zhihao Wang and Liang Zhang and Delai Qiu and Qingsong Liu and Jinsong Su},
title={A Multi-Agent Framework with Automated Decision Rule Optimization for Cross-Domain Misinformation Detection},
year={2025},
month={March},
cdate={1740787200000},
journal={CoRR},
volume={abs/2503.23329},
url={https://doi.org/10.48550/arXiv.2503.23329}
}