Roberta Base Formality Ranker

🚀 English Sentence Formality Prediction Model

This model is trained to predict whether English sentences are formal or informal, offering a practical solution for text style analysis.

✨ Features

• Trained to predict the formality of English sentences.
• Utilizes roberta-base as the base model.
• Trained on multiple datasets including GYAFC and an online formality corpus.
• Applies data augmentation techniques to reduce over - reliance on punctuation and capitalization.

📦 Installation

No installation steps are provided in the original document.

💻 Usage Examples

No code examples are provided in the original document.

📚 Documentation

Model Details

• Base model: roberta-base
• Datasets:
  • GYAFC from Rao and Tetreault, 2018.
  • online formality corpus from Pavlick and Tetreault, 2016.
• Data augmentation: Changing texts to upper or lower case; removing all punctuation, adding dot at the end of a sentence. This is to prevent the model from over - relying on punctuation and capitalization.
• Loss: Binary classification (on GYAFC), in - batch ranking (on PT data).

Performance Metrics on Test Data

Performance on GYAFC

Performance on P&T subset

🔧 Technical Details

The model uses a binary classification loss on the GYAFC dataset and in - batch ranking loss on the PT data. Data augmentation techniques are applied to make the model less dependent on punctuation and capitalization, enabling it to focus more on other text features.

📄 License

This model is released under the Creative Commons Attribution - NonCommercial - ShareAlike 4.0 International License.

📚 Citation

If you are using the model in your research, please cite the following paper where it was introduced:

@InProceedings{10.1007/978-3-031-35320-8_4,
  author="Babakov, Nikolay
  and Dale, David
  and Gusev, Ilya
  and Krotova, Irina
  and Panchenko, Alexander",
  editor="M{\'e}tais, Elisabeth
  and Meziane, Farid
  and Sugumaran, Vijayan
  and Manning, Warren
  and Reiff-Marganiec, Stephan",
  title="Don't Lose the Message While Paraphrasing: A Study on Content Preserving Style Transfer",
  booktitle="Natural Language Processing and Information Systems",
  year="2023",
  publisher="Springer Nature Switzerland",
  address="Cham",
  pages="47--61",
  isbn="978-3-031-35320-8"
}