The first report proposed the architecture of the collaborative working environment ACE . ACE is the multi-agent system that supports dynamic work processes by managing messages. This second report supports community formation by adding the expert recommending system, Expert Recommender , to ACE. In Expert Recommender , we use a hybrid information filtering approach that utilizes both of the content- based filtering and the collaborative filtering techniques that have been studied in several Recommender Systems. And we extend the expert representation model to address difficulties of representing multi-skilled experts. By integrating those approaches, Expert Recommender shows its efficiency through test scenarios.