A machine learning project built in Python that acts as a smart wine recommender. Instead of just filtering by tags, it reads and understands over 130,000 text reviews using AI, allowing for highly personalized, multi-preference suggestions.

Features

• AI Text Embeddings: Uses OpenAI's text-embedding-3-small to understand the semantic meaning behind wine reviews.
• Lightning-Fast Search: Utilizes FAISS (Facebook AI Similarity Search) to quickly find the best matches out of a massive dataset.
• Custom Weighting: Balances text descriptions with tabular data (like price or age) so users get exactly what they are looking for.
• Clean Data: Includes a robust data preprocessing pipeline that uses regression models to fill in missing dataset values.

How to run this project:

Install Git LFS for embeddings

On Linux

sudo apt update
sudo apt install git-lfs
git lfs install
git lfs pull

Create venv:

python3 -m venv .venv

Activate venv:

On Linux

source .venv/bin/activate

On Windows

.\.venv\Scripts\Activate

Install ipykernel

pip install ipykernel

Register kernel

python -m ipykernel install --user --name ml-project

Install python libraries

pip install -r requirements.txt