Learning to Evolve Procedural Content in Games Using Cultural Algorithms
Sarah Saad, Thomas Palazzolo, Chencheng Zhang, Robert G. Reynolds, Ashley Lemke, John O' Shea, Cailen M. O'Shea · 2022
The procedural content of a virtual reality system is a key contributor to its success. However, it is often the case that the content needs to be adjusted. This may take place for a variety of reasons. This research investigates the possibility of the use of machine learning technology to facilitate the modification of a games content. Here, the Deep Dive system was designed originally to predict ancient site locations. Recently it was repurposed to be used as an educational tool to facilitate aspects of STEM education. This required the modification of the content to support this novel use. An evolutionary learning algorithm, Cultural Algorithm, is employed to facilitate the addition of the new content required for the educational application.