Filters which select items for individual users based upon content suffer from several limitations. The items being filtered must be amenable to parsing by a computer. Furthermore, Content-Based Filters posses no inherent method for serendipitous exploration of the information space.
This thesis proposes Social Information Filtering (SF), a general, a novel approach to information filtering. SF systems filter items based upon other users whose tastes are similar to your own. SF overcomes the limitations of Content-Based Filters.
An implementation employing Social Information techniques is presented. Ringo is a music recommendation system accessible to users via electronic mail. Users rate musical artists and then are able to receive recommendations for further listening.
Several filtering schemes are described, analyzed and compared. Experimental results demonstrate the capabilities of SF and its potential for immediate application.