A logic of emotions for intelligent agents

This paper formalizes a well-known psychological model of emotions in an agent specification language. This is done by introducing a logical language and its semantics that are used to specify an agent model in terms of mental attitudes including emotions. We show that our formalization renders a number of intuitive and plausible properties of emotions. We also show how this formalization can be used to specify the effect of emotions on an agent’s decision making process. Ultimately, the emotions in this model function as heuristics as they constrain an agent’s model.