MovieLens Recommendation System

This project showcases a complex recommendation system developed for a media streaming platform (inspired by Netflix) using the MovieLens 1B synthetic dataset. The system adopts a hybrid approach that combines collaborative filtering, content‑based filtering, and graph‑based recommendation to deliver personalized movie suggestions.

Project Structure

├── LICENSE <- Open‑source license (if selected) │ ├── Makefile <- Makefile with convenient commands such as make data or make train │ ├── README.md <- Top‑level README for developers using this project. │ ├── data │ │ │   ├── external <- Data from third‑party sources. │ │ │   ├── interim <- Converted intermediate data. │ │ │   ├── processed <- Final curated dataset for modeling. │ │ │   └── raw <- Immutable raw data dump. │ ├── environment.yml <- Requirements file for reproducing the analysis environment, e.g., generated via pip freeze > requirements.txt │ ├── models <- Trained serialized models, model predictions, or model summaries │ ├── notebooks <- Jupyter notebooks. Naming convention: number (for ordering), author initial, and a brief - separated description, e.g., 1.0-jqp-initial-data-exploration │ └── src <- Source code for this project │ ├── init.py <- Makes src a Python module │ ├── dataset.py <- Scripts for downloading or generating data │ ├── features.py <- Code for creating modeling features │ ├── modeling │ │ │   ├── init.py │ │ │   ├── predict.py <- Code for running model inference with a trained model │ │ │   └── train.py <- Code for training a model │ └── plots.py <- Code for creating visualizations