SDG Classifier OSDG
| Property | Value |
|---|---|
| Model Type | Multi-class Classification |
| Accuracy | 89.73% |
| CO2 Emissions | 0.065 grams |
| Model URL | HuggingFace/jonas/sdg_classifier_osdg |
What is sdg_classifier_osdg?
The sdg_classifier_osdg is a specialized machine learning model designed to classify text according to the first 15 United Nations Sustainable Development Goals (SDGs). Developed by Jonas and powered by the OSDG.ai community data, this model demonstrates exceptional performance with a macro F1 score of 0.85 and overall accuracy of 89.73%.
Implementation Details
The model is optimized for processing short text passages (approximately 100 words) and implements a multi-class classification approach. It's available through both REST API and Python interfaces, utilizing the Hugging Face Transformers library. An improved version using fine-tuned RoBERTa architecture is also available.
- Exceptional accuracy metrics (Macro Precision: 86.94%, Macro Recall: 84.05%)
- Environmental consciousness with minimal CO2 emissions (0.065g)
- Streamlined integration through HuggingFace's infrastructure
Core Capabilities
- Text classification across 15 SDG categories
- Optimized for short paragraph analysis
- High-precision multi-class predictions
- Easy deployment through REST API or Python
Frequently Asked Questions
Q: What makes this model unique?
This model specializes in SDG classification with remarkably high accuracy while maintaining a minimal environmental footprint. Its optimization for short text analysis makes it particularly effective for paragraph-level SDG classification tasks.
Q: What are the recommended use cases?
The model is ideal for analyzing short text passages (around 100 words) against UN Sustainable Development Goals, making it perfect for document classification, research analysis, and policy alignment verification.
