Creating Models of Real-World Communities with ReferralWeb

Henry Kautz and Bart Selman

Most recommender systems generate recommendations based on anonymous opinions. However, when a person is making critical decisions about his life or his business, he wants the advice of a trusted expert. If he does not personally know such an expert, then he wants a personal reference to such an expert via a series of trusted friends and colleagues. ReferralWeb models social networks of people, and lets users find chains of referrals to experts on arbitrary topics. Furthermore, ReferralWeb creates these models by mining information from publicly available documents on the WWW. This allows access to large, real-world communities, without any burden on the members of the communities.


This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.