Balancing Consistency and Plot Structure in Computational Storytelling

With technological advancements, much effort has been put to develop a computational story generator. One of the prime challenges for computational story generators is balancing the consistency and structure of the plot. Although a few attempts have been made to address this challenge, there is still a lot to desire. In this paper, we propose a computational story generator, based on Makebelieve, that while allowing authors to define plot points in the story, offers consistency with a controllable amount of tolerance. We use Directed Causal Graph and dependency matrix along with planning to generate the events. We then check the fitness of the plan by considering the consistency of character personality throughout the plan. With the example stories, we discuss how the proposed approach affects the plot to ensure consistency along with plot structure.