ml-fuzzy-matching

📂 Files

• eda_cleaning.ipynb: Data loading, eda, cleaning, finalizing column.
• utils.py
• data.py: Load and prepare training data (positive + negative pairs).
• features.py: Extract similarity features from string pairs.
• model.py: Train and evaluate LR and XGBoost Model. and depending on performances save the better model.
• evaluate.py: The cli driver for comparison
• match.py: Inference: take a user query and return best match
• config.py: Centralized constants (thresholds, paths, etc.).
• demo.ipynb: notebook to test everything end-to-end and essentially serves as a demo for others to see how to work with this git repo.

💻 Built with

• Python
• Scikit-Learn
• XGBoost

Correct Working Pipeline:

1. evaluate.py
2. demo.ipynb