CineTrack

A FastAPI movie discovery app with local search, likes/comments, and ML-based recommendations.

Features

• Login and signup with JWT cookies
• Browse movies by category
• Search by title (local CSV + DB)
• Like and comment on movies
• Similar-movie recommendations from precomputed PKLs

Requirements

• Python 3.10+
• pip install -r requirements.txt

Run Locally

uvicorn app:app --reload

Then open:

• http://localhost:8000/login

Recommender Model

The app uses PKL files in model_data/:

• movies.pkl
• similarity.pkl

To regenerate using ml_model.py (uses movies + credits datasets):

python3 ml_model.py

Datasets expected in project root:

• tmdb_5000_movies.csv (or .zip)
• tmdb_5000_credits.csv (or .zip)

Project Structure

• app.py — FastAPI routes
• movie_service.py — core business logic
• templates/ — HTML pages
• styles/ — CSS styles
• scripts/ — frontend JS
• model_data/ — ML PKLs