Adaptive Learning in Mobile Serious Games: A Personalized Approach Using AI for General Knowledge Quizzes
Michail Tselepatiotis, Efthimios Alepis · 2024
This paper introduces a mobile serious game on general knowledge quizzes. It is designed as a zombie shooting game where players must shoot the zombie with the correct answer to each question. The game integrates machine learning and artificial intelligence to tailor the learning and gameplay experience to the player's skill level and performance. We employ a K-nearest neighbors (k-NN) model, trained on data from sample users, to predict in-game question performance for new players and adjust the learning experience by balancing the distribution of question categories based on predicted player performance. Additionally, Fuzzy Logic is used to dynamically adjust zombie behavior, making it more aggressive for skilled players and simpler for less skilled players. This approach provides a dynamic and engaging gameplay experience that adapts to the player's abilities, offering a personalized learning journey while maintaining challenge and immersion. Experimental results and player feedback highlight the system's effectiveness and usability in boosting player engagement and learning outcomes, though there remains potential for further improvement.