ANDROID APPLICATION FOR AUTOMATIC FILM SELECTION ACCORDING TO USER'S PREFERENCE

Abstract

The work outlines the development of a movie and TV show recommendation application, integrating with services like IMDb and TMDB, OMDb. It uses Java and Android Architecture Components, with Jsoup for HTML data processing. The app combines content-based and collaborative filtering for personalized recommendations. Collaborative filtering offers diverse suggestions but faces issues like the "cold start" problem, whereas content-based filtering focuses on user-preferred characteristics but may lack variety. To enhance recommendations, Singular Value Decomposition (SVD) can be employed, reducing data dimensionality and revealing hidden relationships, though it has computational limitations.