An AI-Driven Content-Based Recommendation Framework for Personalized Movie Discovery

Eleena Jena

Department of Master of Computer Applications

GIFT Autonomous, Bhubaneswar, Odisha, India, eleena2024@gift.edu.in

Prajyakta Patra

Department of Master of Computer Applications

GIFT Autonomous, Bhubaneswar, Odisha, India, prajyakta2024@gift.edu.in

Shibani Bardhan

Department of Master of Computer Applications

GIFT Autonomous, Bhubaneswar, Odisha, India, shibani2024@gift.edu.in

Dr. Pratyush R Mohapatra

Associate Professor, Associate Dean Academics

GIFT Autonomous, Bhubaneswar, Odisha, India, assodeanacad@gift.edu.in

  Abstract— The rapid growth of online streaming platforms has significantly increased the availability of digital entertainment content. Although users now have access to thousands of movies across different genres and languages, discovering relevant content according to individual preferences has become increasingly difficult. Traditional search-based approaches are inefficient because they require users to manually browse large collections of movies.

This research presents the design and implementation of an Intelligent Movie Recommendation System developed using Machine Learning and Content-Based Filtering techniques. The proposed system recommends movies by analyzing metadata attributes such as genres, cast information, keywords, and movie descriptions. Count Vectorization and Cosine Similarity algorithms are applied to compute similarity scores and generate personalized recommendations efficiently.

The system integrates Flask-based backend services, MongoDB database management, and TMDB API integration for dynamic retrieval of movie posters and movie details. Additional functionalities including OTP-based email verification, user authentication, favorites management, search history tracking, and responsive user interface design enhance both usability and security.

Experimental analysis demonstrates that the proposed recommendation system successfully generates relevant movie suggestions with minimal response time while maintaining smooth interaction and efficient recommendation processing. The system provides an intelligent and scalable solution for personalized movie discovery.

    Keywords— Machine Learning, Recommendation System, Content-Based Filtering, Cosine Similarity, Flask Framework, MongoDB, TMDB API, Artificial Intelligence.