README.md

---
license: lgpl
language:
- en
- de
- es
- ca
metrics:
- accuracy
pipeline_tag: text-classification
tags:
- stance
- classification
- pytorch
- multilingual
---

## Model Description

<!-- Provide a longer summary of what this model is. -->

- **Developed by:** Cristina España-Bonet
- **Model type:** Binary stance classifier on top of XLM-RoBERTa
- **Language(s) (NLP):** English, German and Spanish
- **License:** LGPL
- **Finetuned from model:**  XLM-RoBERTa Large

## Model Sources

<!-- Provide the basic links for the model. -->

- **Repository:** https://github.com/cristinae/docTransformer
- **Data:** https://zenodo.org/records/8417761
- **Paper:** https://aclanthology.org/2023.findings-emnlp.787/


## Direct Use

Determine the political stance of a (newspaper) article. Binary classification: left vs. right stance

#### Evaluation

```srun --ntasks 1 --gpus-per-task 1 python -u docClassifier.py --task evaluation -f ./ -o politicalStanceLvsR_en.bin  --test_dataset your.test```

#### Classification

```srun --ntasks 1 --gpus-per-task 1 python -u docClassifier.py --task classification -f ./ -o politicalStanceLvsR_en.bin  --test_dataset your.test```

## Citation [optional]

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**
```
@inproceedings{espana-bonet:2023,
    title = "Multilingual Coarse Political Stance Classification of Media. The Editorial Line of a {C}hat{GPT} and Bard Newspaper",
    author = "Espa{\~n}a-Bonet, Cristina",
    editor = "Bouamor, Houda and Pino, Juan and Bali, Kalika",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2023",
    month = dec,
    year = "2023",
    address = "Singapore",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2023.findings-emnlp.787",
    doi = "10.18653/v1/2023.findings-emnlp.787",
    pages = "11757--11777"
   }
```

**APA:**

España-Bonet, Cristina. (2023, December). Multilingual Coarse Political Stance Classification of Media. The Editorial Line of a ChatGPT and Bard Newspaper. In Findings of the Association for Computational Linguistics: EMNLP 2023 (pp. 11757-11777).