Evolving story and character generation for role-playing games

In this paper the authors present a new approach by which a random story with random characters can be generated by the use of Genetic Algorithms. Stories are a collection of several tasks done by the characters in a certain sequence. If every task is taken as an independent entity, random distribution of these tasks can create a new story. We can then use Genetic Algorithms to create the characters required by those stories. Through user interaction, we can dictate what kind of characters and stories should be created and the computer slowly converges to creating stories according to the users liking. With the implementation of this methodology, we can ensure that no two plays of the same game should give the user the same experience. With user input we can ensure that the Genetic Algorithm learns with every play-through thus creating stories and characters that are more according to the liking of the user.