In-Vehicle Infotainment Systems: Using Bayesian Networks to Model Cognitive Selection of Music Genres

In the world of transportation, the ceaseless evolution of Information and Communication Technologies (ICT) is reflected on their migration towards In-Vehicle-Infotainment (IVI) systems, which are characterized by innovative applications and services. Such systems aim to support drivers/passengers with a varying set of functions, enhancing the quality of driving through entertainment. The goal of this study is to introduce a novel IVI cognitive functionality that automatically and dynamically proposes the optimal music genre to the drivers/users when they want to make a certain journey with their owned (or shared) vehicles. The proposed recommender functionality utilizes in an automated manner (i) drivers/users’ profile data and current situation, (ii) drivers/users’ personal preferences, (iii) external environment information obtained from sensor measurements, and (iv) previous knowledge and experience. Knowledge is obtained through the exploitation of Bayesian networking principles in combination with a practical implementation of the Naive-Bayes model. Indicative simulation results showcase the efficiency of the proposed infotainment functionality, in terms of accuracy and speed of convergence, in proactively identifying the optimal music genre and accordingly notifying the drivers/users.