The code related to the paper is below:
RumorMixer: Exploring Echo Chamber Effect and Platform Heterogeneity for Rumor Detection, ECML-PKDD, 2024.
It is a differentiable architecture search framework based on GraphMixer for rumor detection.
To execute this project, it's essential to install the required dependencies. To do so, navigate to the directory containing the requirements.txt
file and execute the following command:
pip install -r requirements.txt
To execute the project, run the following command in your terminal:
Step 1. Run the search process, given different random seeds. (The Twitter16 dataset is used as an example)
python Libs/Exp/search.py
The results are saved in the directory Output/Logs/
, e.g., Search-RumorMixer-Twitter
.
Step 2. Fine tune the searched architectures. You need specify the arch_filename with the resulting filename from Step 1.
python Libs/Exp/finetune.py
Step 2 is a coarse-graind tuning process, and the results are saved in a picklefile in the directory Finetune
.
If you find this project useful and use it in your research, please consider citing our paper.
The code is built on DARTS, one of the most well-known differentiable architecture search methods and SANE, a differentiable architecture search for graph neural network (GNN).
If you have any questions about this project, you can open issues, thus it can help more people who are interested in this project. We will reply to your issues as soon as possible.