Procedural constrained story generation based on Propp's and Fabula models

Abstract

Using narrative to increase engagement in serious games where the user is required to repeatedly interact with the same game for a long period of time, as in rehabilitative exer-games, could help in reducing the boredom inherent to the task. However, producing enough narrative content to cover long rehabilitation periods could require much effort from the authorial point of view. Having a tool that, given a wide enough knowledge base, can virtually produce an infinite set of stories, could greatly help in addressing this issue. For this reason, we explored the field of automatic story generation, designed and developed a multi-layered tool capable of automating the creation process of stories. The tool allows the user to define constraints at setup-time to ensure that a given set of events will appear in the generated story plot. These events may be mapped to specific serious games, making the resulting stories suitable, for instance, for integration in exer-games based rehabilitation platforms.