This paper reviews the literature surrounding an information filtering technique, collaborative information filtering, which supports the discovery of resources in a way that is sensitive to the context of users. Moreover, via statistical clustering techniques, the system supports automated, personalized filtering and recommendation of relevant resources and like-minded users for particular user communities. The paper also describes an educational implementation of this approach, called Altered Vista, and presents results from a 3-month trial use of the system, aimed at evaluating the educational effectiveness and usefulness of the approach.