Mind maps as behavior controllers for virtual characters

We propose a new mind map interface method for controlling virtual character's artificial intelligence and behavior in games. Commonly used techniques, such as scripting, require an extensive analysis and discrimination of all the possible behaviors that are triggered by the stimuli received by the virtual character. Scripting is also subject to strict syntax rules that may be hard to comprehend by non-expert users. Instead, our method can be easily created by users that do not have any technical background, since they graphically represent the natural process of organizing information in the human brain. Our mind map interface method follows a behavior-based architecture combined with an emotional depth module to control the character's behaviors individually in a game. We implemented a graph-based visual editor to ease the definition of the mind map nodes interactively. We also show how mind maps were implemented in the LIFEisGAME (LearnIng of Facial Expressions usIng Serious GAMEs) project as a proof of concept.