Design And Implementation Of Personalized Interactive System For Smart TV

In recent years, with the penetration of Internet-related technologies, smart TVhas been rapid development, and market competition is also mounting up. Unlikethe traditional TV, smart TV will not only provide users with rich applications, alsoneed to provide users rich interactive experience. With new remote control for smartTV designed and launched, making prossible to identify user identity via remotecontrol behavior, also can through the new remote control to complete the newinteractive experience.Based on the new remote control equips orientation sensor and gravity sensor,and Smart TV User Identification System, this paper design and imple mentation ofprogram recommendation, and the use of remote control mouse functions, these twofunctions together to complete the personalization interactive.The mouse control function, through remote mouse interaction subsystems, bysensor service and mouse service two modules to complete, complete sensor dataacquisition, and mouse movement calculation. With supporting of sensors byAndroid system, remote mouse interaction subsystems complete the acquisition ofsensor data, and through AIDL interface definition language communicationtechnicial to achieve data sharing. Meanwhile, through the changes od the sensors’sdata, constructint a simulate moving track algorithms, and propose two steps ofhigh-pass and little-pass data processing method for the inversion of remote control,complete the mouse pointer movement operation.The recommendation function, through user behavior subsystems, and by dataacquisition, interest modeling, and behavioral analysis three modules to complete.Including complete the resources acquisition, status monitoring, browsing datacollection, information processing, user interest model, user interest calculations,user interest recommend functions. Through message callback and notificationmodel, the user behavior framework subsystems get system status and resource bySmart TV frame system, and through the regular expression matching technique tocomplete the information process. Then, by c onstructing the interest attenuationmodel, and the relation degree model, to calculate the user interest degree andrecommended coefficient, complete the interest recommend for users. And througha probabilistic representation method, prevent a bad interaction experience causedby system frequently recommended when user frequently in operation.After the completion of system imp lementation, for the various functions of thesystem gives the test, also the system performance assessment test. The test results generally in line with expectations, meet the system demand.