CineMatch is a dynamic movie recommendation platform tailored for movie enthusiasts seeking to explore new films, engage with a community of like-minded individuals, and enjoy a personalized movie-watching experience. Our platform not only simplifies the discovery of movies but also fosters community engagement through user interactions, comments, and filmbuddy matchmaking.
- User Authentication: Secure login and registration system to manage your profile.
- Movie Discovery: Extensive search functionalities to explore a vast database of films.
- Personalized Recommendations: Custom movie suggestions tailored to your individual preferences.
- Community Engagement: Connect with other users, share your thoughts on movies, and follow users with similar tastes.
- List Management: Craft and curate custom movie lists to track your watched films and plan your watchlist.
- Filmbuddy Matchmaking: Discover and connect with users who share your movie preferences.
- User Profiles: Access detailed movie-watching statistics and customize your user profile.
Our approach uses hybrid models to deliver unique movie recommendations:
- Collaborative Filtering: Using the SVD algorithm, we predict how much you will enjoy unseen movies based on your rating history and that of others with similar tastes.
- Content-Based Filtering: For users new to the platform or those without sufficient rating history, we recommend popular movies based on content similarity and vote averages.
- Hybrid Recommendations: By blending collaborative and content-based suggestions, we ensure a richly personalized selection of movie recommendations.
- Random Popular Movies: As a fallback, new users without any ratings are presented with random popular movies, guaranteeing quality suggestions for everyone.
- Cosine Similarity: We employ cosine similarity to identify users with similar movie rating profiles, enabling us to recommend movies based on user similarities.
- Dynamic Updates: Our system integrates with a PostgreSQL database for real-time data retrieval, ensuring that recommendations are always current and relevant.
- Personalized Matching: By identifying top-N similar users, we offer personalized movie suggestions, enhancing the user experience through tailored recommendations.
- Semantic Analysis: Utilizing a precomputed cosine similarity matrix with BERT natural language processing model, our system identifies movies with closely related genres, overview, director etc... offering recommendations that truly resonate with your preferences.
