In recent years, a few group recommender systems have been designed (Masthoff, 2002; McCarthy & Anagnost, 1998; O’Connor et al., 2001). All of these systems assumed that the input of the system is comprised of items’ ratings given by individuals, and they obtain the group recommendations by combining or aggregating the individual recommendations of the members in the group. These previous methods can be categorized into the following three approaches: