Etudiant : EL-AMARTY mohamed

Filière : LF Sciences Mathématiques et Informatique

Encadrant : Pr. NFAOUI EL HABIB

Annèe : 2025

Résumé : In today's digital world, users face an overwhelming amount of content, making it increasingly difficult to find movies that match their preferences. This final-year project aims to design and develop an intelligent movie recommendation system that provides a personalized viewing experience. Our solution is based on a hybrid approach that combines three cutting-edge techniques: Matrix Factorization (SVD) to analyze user ratings and identify hidden preference patterns. Large Language Models (LLM) with FAISS to enable semantic search through natural language queries. A Deep Learning model (VGG16) to analyze movie posters and predict visual preferences. The system is built using the Django framework for the backend, with a modern user interface developed in React. Each recommendation technique is accessible through a separate view, allowing users to choose the recommendation mode that best suits their needs. Additional features such as user registration, login, and recommendation history are also included. This project provides a robust and scalable solution to common recommendation system challenges such as cold start, data sparsity, and lack of contextual understanding, showcasing the power of modern AI in delivering intelligent and user-centric applications.