📢 Release Notes

BibexPy Hydrogen v1.1.0 - Hybrid Enrichment Release

🚀 New Features & Enhancements

Improved Machine Learning-Based Enrichment (Still Experimental)
- ML-based enrichment is now more structured and precise, leveraging TF-IDF + RandomForest models.
- Enhanced handling of training and empty records for improved prediction accuracy.
New Hybrid Enrichment Option: API + ML Enrichment
- A new option combines API-based enrichment with ML-based enrichment, providing more flexibility and accuracy in metadata enhancement.
Updated Field Merging Algorithms
- Optimized merging logic to improve data consistency and processing efficiency.
Resolved macOS Compatibility Issues for API Enrichment
- API-based enrichment now operates correctly on macOS systems without compatibility issues.

Optimized Processing Speed
- Various backend optimizations have been implemented to improve efficiency and reduce execution time.
Enhanced Logging and Statistics
- More detailed logging for ML model training, record distributions, and enrichment processes.

General Error Fixes & Stability Improvements
- Several issues related to metadata enrichment and ML-based processing have been resolved.
- Improved robustness of enrichment workflows to handle edge cases more effectively.

Let us know if you encounter any further issues! 🚀