deberta-large-formality-ranker
| Property | Value |
|---|---|
| Base Architecture | DeBERTa-large |
| License | OpenRAIL++ |
| Paper | Detecting Text Formality (RANLP 2023) |
| Accuracy | 87.8% |
What is deberta-large-formality-ranker?
This is a specialized model designed for detecting text formality levels, based on the DeBERTa-large architecture and fine-tuned on the GYAFC (Graded Yahoo Answers Formality Corpus) dataset. It represents the state-of-the-art in transformer-based approaches for formality classification, achieving impressive accuracy of 87.8% and balanced F1-scores for both formal (89.0) and informal (86.1) text classification.
Implementation Details
The model builds upon the DeBERTa-large architecture and has been specifically optimized for binary formality classification tasks. It outperforms other transformer-based models including DistilBERT-cased, and shows competitive results against character-based BiLSTM approaches.
- Excellent performance metrics: 87.8% accuracy, 89.0 F1-score for formal text, 86.1 F1-score for informal text
- Easy integration with the Transformers library
- Robust evaluation across multiple text classification approaches
Core Capabilities
- Binary classification of text formality (formal vs informal)
- Balanced performance across both formal and informal categories
- Suitable for both academic and industrial applications
- Cross-lingual formality detection potential
Frequently Asked Questions
Q: What makes this model unique?
This model represents the current state-of-the-art in transformer-based formality detection, offering superior performance metrics and balanced classification capabilities for both formal and informal text. It's particularly notable for its robust performance compared to other approaches in the field.
Q: What are the recommended use cases?
The model is ideal for applications requiring automatic formality detection in text, such as content moderation, style analysis, automated writing assistance, and academic research in linguistics. It's particularly useful in scenarios where distinguishing between formal and informal writing styles is crucial.
