Composition of probabilistic evaluations of preferences: a case of criteria applied to isolated and clustered options.

This article applies a transformation into probabilities of preference, to rank papers submitted to the XLI Brazilian Symposium of Operational Research. Some 42 papers were divided into sets according to their main topics of interest and measures for 12 attributes were considered, for each paper separately, and also for sets of papers on the same topic, based on the referees' evaluations. Four alternative forms of composition of the probabilistic preferences are illustrated. [ABSTRACT FROM AUTHOR]