Graph-Based Movie Recommender 🎬🔗

This project implements a graph-based movie recommendation system using link prediction algorithms. The system predicts which movies a user might like based on past interactions using graph similarity metrics.

📁 Files

• main.ipynb: Jupyter Notebook containing the full implementation.
• dataset.csv: Input dataset of user-movie interactions.
• train_dataset.csv, test_dataset.csv: Data splits used for training and testing the model.

📌 Project Description

The core idea is to model the user-movie interactions as a bipartite graph, where:

• Nodes represent users and movies.
• Edges represent interactions (e.g., watched or rated).

We apply link prediction techniques to estimate the likelihood of future user-movie interactions, effectively recommending movies that a user is likely to enjoy.

🧠 Techniques Used

• Graph Construction using NetworkX
• Link Prediction Algorithms:
  • Jaccard Coefficient
  • Adamic-Adar Index
  • Preferential Attachment
  • Resource Allocation Index
• Train/Test splitting per user
• Visualization of sample graphs

⚙️ How to Run

1. Clone the repository:

   git clone https://github.com/your-username/movie-recommender-graph.git
   cd movie-recommender-graph