Research And Implementation Of Film And TV Program Recommendation System For Interactive TV

The information on the Internet increases with the development of the Internet and broadcast networks, and the Internet itself has been the greatest information bank of the world especially with the rapid development of the technology of WEB2.0, which brings easy access as well as information expansion. The tri-networks integration, the strategic objective of the development of China’s information industry, is making a difference to our life. With the policy of Accelerate the Development of Tri-networks Integration issued by the State Council in2010, the Next-generation Broadcasting in recent years gains rapid development, making the set-top boxes on interactive demand into our daily life and more videos on demand for users, thus the TV users enjoy such all-embracing service experience as interconnection, personalized search and recommendation.With the soaring number of videos in digital TV networks for users, users would become at a loss to choose from the hundreds and thousands of videos for the search engine is only partially helpful. Therefore, the research focus has been moved to the interactive set-focused personalized recommendation system.One important function of interactive TV is the search and on-demand service of films and TV programs. With more and more such programs and more complex contents, users are also confronting confusion in choosing, however, the video recommendation technology can be an ideal solution. Based on the users’ play records, this technology can analyze users’ preferences, and then generate a list for recommending in the light of the information of interactive TVs. This thesis mainly studies the recommendation technology of films and TV programs in the on-demand system of interactive TVs, and achieves a corresponding personalized recommendation system. By analyzing the correlation of films and programs and the play records, this system can dig for information about users’ similarities and preferences, thus making the recommendation personalized and the troublesome choosing relieved, and finally achieving better user experience.