Andrew W. Wicker, Jon Doyle
Comparisons of similarity or dissimilarity between systems of preferences over multiple attributes play important roles in interest matching, social networking, collaborative filtering, and personalization. We develop metrics over preferences represented compactly by conditional preference networks (CP-networks) and their variants. Our metrics exhibit intuitive properties and support efficient (polynomial-time) algorithms for computing similarities.
Subjects: 15. Problem Solving; 15.2 Constraint Satisfaction
Submitted: May 5, 2008