A Reinforcement Learning System for Adaptive Gamification and Hexad User Profile Tracking

Gamification has become a widely used user moti-vation approach in modern apps, since it can adapt to the user's characteristics, such as the Hexad user type, to achieve better performance. Despite various machine learning techniques being employed and tested in gamification research, reinforcement learning has seen little use in gamification in order to assess its usefulness. Thus, in this paper we designed a customizable adaptive gamification reinforcement learning system that uses the user's profile to change the gamification elements shown to them and to observe the changes their profile undergoes over time. A prototype of the designed system was created using the design conventions of the Gymnasium library (formerly OpenAI Gym). The reinforcement learning system underwent preliminary tests regarding its accuracy and profile tracking using user answer simulation. Results indicate that the system functions satisfactorily in various use cases and that its accuracy is high when the system closely tracks the user's Hexad profile.