Deep Multi-Task Model for Sarcasm Detection and Sentiment Analysis in Arabic Language

Requirement

1. pytorch
2. transformers
3. scikitlearn
4. pandas
5. barbar

Model Training

python train_model.py [args]

Options:

• lm_pretrained : pretrained BERT model (MARBERT, ARBERT, AraBERT, ...)
• lr : learning rate
• batch_size : batch size
• epochs : number of epochs
• lr_mult : Classifier learning rate multiplier

Model evaluation/testing

python eval_model.py --lm_pretrained [value] --batch_size [value]

Citing this work

If you are using this source code please use the following citation to reference this work:

@inproceedings{el-mahdaouy-etal-2021-deep,
    title = "Deep Multi-Task Model for Sarcasm Detection and Sentiment Analysis in {A}rabic Language",
    author = "El Mahdaouy, Abdelkader  and
      El Mekki, Abdellah  and
      Essefar, Kabil  and
      El Mamoun, Nabil  and
      Berrada, Ismail  and
      Khoumsi, Ahmed",
    booktitle = "Proceedings of the Sixth Arabic Natural Language Processing Workshop",
    month = apr,
    year = "2021",
    address = "Kyiv, Ukraine (Virtual)",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/2021.wanlp-1.42",
    pages = "334--339",
}