The value of information as a verification and regret-preventing mechanism in algorithmic search environments

Information is a fundamental asset in any organization and as such is assigned a value reflecting its importance. We define information to be valuable if it either confirms the choice that a decision maker (DM) would make based on the information acquired on a given alternative or prevents him from making a regrettable choice, or both. We introduce a novel information acquisition algorithm where the value of information is used as a verification and regret-prevention mechanism determining the behavior of the DM. The proposed algorithm shows how the incentives of the DM to continue acquiring information on a given alternative are determined by his attitude toward regret and the relative spread exhibited by the domains on which the characteristics of the alternative are defined. Moreover, our model proves the existence of relative spread scenarios where indecision arises, leading the DM to behave randomly. The generality and flexibility of the model allows to easily develop extensions and applications to decision theory, psychology, economics and operational research.