Smart marketing in Brazilian digital TV system through a recommendation ads

Abstract

With the implementation of the Brazilian Digital TV System (SBTVD), starts a range of new opportunities and possibilities both for viewer as for TV stations. For the viewers, they will have an immense amount of channels, programs and interactive advertisements. For TV stations, it increases the possibility of advertising in new media. Within this framework, the opportunity arises for a recommendation system for applications and interactivity portals. This dissertation presents a proposal of advertising personalization into applications and portals of digital TV environment in order to bring a better experience to the viewer, a new form of income for the broadcasters and also a greater acceptance of specialized products for use. This work develops an application for interactive Digital TV called Smart Marketing, capable of capturing viewer navigation data through both implicit and explicit means by performing customized advertising from the process of knowledge discovery. Developed from AstroTV middleware, compatible with the Brazilian specification, its application was evaluated by means of experiment that used varied user profiles, applying into the generated database the process of knowledge discovery, which used tasks of classification and grouping. The results indicated the quality of the recommendation generated by Smart Marketing.