Context-Aware Movie Recommender System

This project implements a context-aware movie recommender system using the MovieLens 20M dataset. It uses Context-Aware Matrix Factorization (CAMF) to provide personalized movie recommendations based on user ID, time of day, and day of the week.

Project Structure

context_aware_recommender/
│
├── docs/
│   ├── ProjectCharter.md
│   └── DeploymentInstructions.md
│
├── notebooks/
│   ├── DataExploration.ipynb
│   └── EvaluationMetrics.ipynb
│
├── src/
│   ├── data_preparation.py
│   ├── model_training.py
│   └── app.py
│
├── Dockerfile
├── requirements.txt
└── README.md

Getting Started

Prerequisites

• Python 3.9+
• Docker (for deployment)

Installation

1. Clone this repository:

   git clone https://github.com/zeeenoo/Context-Aware-Movie-Recommender-System.git
   cd context-aware-recommender
   

2. Install the required packages:

   pip install -r requirements.txt
   

Usage

1. Data Preparation:

   python src/data_preparation.py
   

2. Model Training:

   python src/model_training.py
   

3. Run the Streamlit app:

   streamlit run src/app.py
   

Deployment

To deploy the application using Docker, follow the instructions in docs/DeploymentInstructions.md.

Dataset

This project uses the the MovieLens 20M dataset, which is loaded directly from Kaggle Datasets. The dataset includes movies rating, user information, and timestamps, allowing for context-aware recommendations.

Acknowledgments

• the MovieLens 20M dataset provided by kaggle Datasets
• Inspired by research on context-aware recommender systems
• Streamlit for providing an easy-to-use framework for building data applications

For more information on the project goals, timeline, and success criteria, please refer to the ProjectCharter.md file in the docs directory.