Collaborative Filtering as means to reduce information overload

One fundamental problem that exists with the World Wide Web is the difficulty involved in locating relevant material given the enormous quantities of information available. Existing systems try to tackle this problem by using spiders to locate, and indexes to provide access to, these resources. Although useful, there are a number of flaws associated with this approach. We describe an approach using collaborative (or social) filtering which attempts to overcome these flaws. Collaborative filtering automates the informal social system where one person recommends material to another, by recognising shared interests and recommending material based on shared interests.