An affordable and inclusive system to provide interesting contents to DTV using Recommender Systems

Abstract

This paper presents a television scheduling system that aims to aid developing countries in digital television transition by providing content interesting for the people while reducing content acquisition costs. It is based on recommender systems for audiovisual content with special considerations for groups of viewers. The system proposed is fed with content uploaded by the people, properly classified. In a first and basic scenario the system generates four different thematic signals with predefined genres. In the second one, the profile for each user is known and audience is modeled as a group of viewers with known user profiles, allowing its segmentation and channel scheduling according to actual preferences.