Fusion of qualitative preferences with different vocabularies
Publication Type:Conference Paper
Source:Lecture Notes in Artificial Intelligence, Springer-Verlag, Volume 2504, p.113-124 (2002)
We present a method for the aggregation of qualitative preferences. This method is specially indicated when the set of criteria that is expressing the preferences is not homogeneous. That is, criteria have vocabularies with completely different terms, vocabularies with some common terms with different semantics or vocabularies with different granularity. The goal of the fusion process is to obtain a new ordered qualitative criterion expressing the overall preference over a set of alternatives. Our approach uses clustering techniques as an aggregation tool, and the principal components analysis to obtain a ranking of the alternatives.