Systems such as the Evolving Agent, Lyrictime, and INFOSCOPE all suffer from a ``cold-start'' problem in that new users start off with nothing in their profile and must train a profile from scratch. During the training period the system can't effectively filter for the user, and this initial hump may convince many users to stop using the filtering system, or even give up on the information source entirely. A better system would allow new users some type of access to the experiences of current users to help create an initial profile.
A more general problem with systems that rely on user profiles is that the user can become circumscribed by their profile. The profile only selects articles similar to the ones the user has already read, and new areas which might be of interest can be missed completely. Further, if the user tries to explore areas outside the knowledge domain of the profile, the profile can provide little if any filtering ability, so the user will again confront the cold-start problem and be left facing a staggering amount of data to search through manually. To support the exploration of new areas what is needed is a method for tapping the knowledge of users currently well versed in those areas.