Story telling engine base on agent interactions


Comics have been successfully used as a programing language for agents, giving them instructions on how to act and how to do it; If we take that model and reverse it, we can use comics to describe the actions of agents already interacting with each other, thus, creating a storytelling engine that can generate stories dynamically based on the interaction of said agents.

The model for the agent behaviours is based on the emotional model of Hayes Roth. This model is chosen due to the nature of comics themselves. Comics like those found on newspapers and children magazines are funny because their characters behavior depend heavily on emotions and commonly express them in exaggerated ways.

Our project implements this emotion-based model for agent behavior in a way that tells a story in the form of comic strips. For this, the model will be adapted to a discrete space-time form since the actions no longer occur in real time (like in traditional simulations games) but rather in a sequence of frames or panels. This changes are inspired in the analisis of time and space mechanics in comics by Scott McCloud. The emotional model will also be adapted to reflect the rather extreme emotions and responses that characterize cartoon characters.