REPRESENTING AND AGGREGATING PREFERENCES 
USING A STOCHASTIC INTERPRETATION


In this talk, we present a new way of representing and aggregating
preferences using a stochastic interpretation.

Given a (weighted) list of (binary) preference relations on a set of
alternatives, our objective is to aggregate this information into a 
single preference relation.

Introducing a Markov chain where each state corresponds to an alternative
and whose transition probabilities are computed using the list of preference
relations, we rank the alternatives according to the resulting asymptotic 
probability distribution.

We also extend the method to introduce the notion of incomparability and
present a novel graphical scheme to represent alternatives. We demonstrate
various properties of this method, especially in the important case where
the preference relations are semiorders.