Probabilistic Modeling for Context-Aware Service in Smart TV

Smart TV services which have been provided a one-way or simple united content by content business mostly focus on content services. The smart TV contains a variety of functions causes a problem that the user has to learn the functions. To solve this problem, we propose context-aware service in the smart TV using modular Bayesian networks. Bayesian network can be designed by domain knowledge when there is not enough data. It has been emerged as a powerful technique for context-aware services. We verify the proposed method is useful. It is prove by measuring average of performance of service which gives the result 88% and conducting subjective test provide the result of 76.5.