Hybrid Movie Recommendation System

Overview

• This project uses the MovieLens 25M dataset.
• Functions include data processing, exploratory analysis, and movie recommendations using machine learning.

Installation

pip install pandas numpy scikit-learn surprise plotly matplotlib dask[complete]

Data Setup

1. Download MovieLens 25M from MovieLens.
2. Unzip ml-25m.zip in the project root. Ensure CSV files are inside.

Usage

• Run rc_final.ipynb.
• Run a single part at a time.
• Use factor_of_data variable to load only a subset of original MovieLens dataset.
• Follow comments in the notebook for guidance.

Features

• Data merging and preprocessing.
• Exploratory analysis with basic stats and plots.
• Recommendation models using Surprise and Scikit-Learn.
• Output predictions to CSV.

Models

• Models:

1. Popularity based model
2. Content based model
3. Collaborative Filtering
4. Matrix Factorization method
5. Combined model (SVD + CF)
6. Hybrid model

Similarity Metrics

1. Cosine similarity
2. Mean square difference-based similarity.
3. Pearson coefficient (mean-centred cosine similarity)
4. Pearson Baseline (uses global baselines for centring instead of means)

Visualizations

• Generate plots using matplotlib and plotly (if uncommented).

License

• MIT License. See LICENSE file.