Abstract
Consider a decision maker (DM) who must select an alternative to evaluate when using an online recommender engine that displays multiple evaluations from unknown raters regarding the different characteristics of the available alternatives. The evaluations of the raters do not necessarily coincide with those that would be provided by the DM, who must consider the differences existing between the ratings observed and his subjective perception and subsequent potential evaluations. We formalize the incentives of the DM to observe and evaluate an alternative through a function that accounts for these differences in a multi-criteria decision making setting. The resulting perception-based framework is implemented in a data envelopment analysis (DEA) scenario to analyze the effects of perception differentials on the evaluation and ranking behavior of DMs.