(1) IBM T.J. Watson Research Center, 10598 Yorktown Heights, NY, USA;(2) IBM Software Group, NY 10532 Hawthorne, USA;(3) IBM China Research Lab, Beijing, China
Abstract:
TV program ratings directly affect the decisions of advertisers and TV stations program offerings. TV program providers use television ratings reports to offer appropriate programs as well as attract advertisers. In this paper we propose a framework for TV audience estimates and program ratings using a per-object-granularity tracking mechanism for interactive TV content tracking in real time. It can be applied to live TV broadcasting environments as well as recorded broadcasting environments. The major components of the proposed tracking architecture are an interactive TV content creation tool powered by hotlinked video technology, a tracking information delivery framework with live interactive content insertion capability, an interactive TV content receiver, and tracking servers. The proposed mechanism and system in this paper enables a broadcast station or a rating agency to automatically perform tracking functions in real time based on viewer actions as well as on the content of audiovisual programming.
program offerings. TV program providers use television ratings reports to offer appropriate programs as well as attract advertisers. In this paper we propose a framework for TV audience estimates and program ratings using a per-object-granularity tracking mechanism for interactive TV content tracking in real time. It can be applied to live TV broadcasting environments as well as recorded broadcasting environments. The major components of the proposed tracking architecture are an interactive TV content creation tool powered by hotlinked video technology, a tracking information delivery framework with live interactive content insertion capability, an interactive TV content receiver, and tracking servers. The proposed mechanism and system in this paper enables a broadcast station or a rating agency to automatically perform tracking functions in real time based on viewer actions as well as on the content of audiovisual programming.