In order to integrate properly recording services with other streaming functionalities in a DMR (e.g., AppleTV, PS3) we need a way to put live TV and radio events into friendly catalogs. But recordings are based on parameters to be set by the users, such as timings and
channels, and event discovery can be not trivial. Moreover, personalized recommendations strongly depend on the information quality of discovered events.
In this paper, we propose a general collaborative strategy for discovering and recommending live events from recordings with different timings and settings. Then, we present an analysis of collaborative filtering algorithms using data generated by a real digital video and radio recorder.