Conditional-Adversarial-Networks-for-Multi-Domain-Text-Classification

Implementation of "Conditional Adversarial Networks for Multi-Domain Text Classification" in Pytorch (Adapt-NLP at EACL 2020)

Datasets

This folder contains the dataset in the same format as needed by our code.

Requirements:

• Python 3.6
• PyTorch 0.4
• PyTorchNet
• scipy
• tqdm (for progress bar)

Training

All the parameters are set as the same as parameters mentioned in the article. You can use the following commands to the tasks:

MDTC experiments on the Amazon review dataset

cd code/

python train.py --dataset prep-amazon --model mlp

Citation

If you use this code for your research, consider citing:

@article{wu2021conditional,
  title={Conditional Adversarial Networks for Multi-Domain Text Classification},
  author={Wu, Yuan and Inkpen, Diana and El-Roby, Ahmed},
  booktitle={Proceedings of the Second Workshop on Domain Adaptation for NLP},
  pages={16--27},
  year={2021}
}