Tee, Eng Bah, and Song, Insu (2025) Evaluation of Prediction Model for Mobile Educational Games. In: Lecture Notes in Networks and Systems (1441) pp. 115-132. From: ICICT 2025: Tenth International Congress on Information and Communication Technology, 18-21 February 2025, London, UK.

PDF (Published Version) - Published Version Restricted to Repository staff only

 

2

Abstract

Mobile educational games have arisen as a fascinating tool to teach difficult concepts in an interactive and engaging manner. Mobile educational game uses a game type like puzzle, strategy, role-playing, and so forth to drive the education of learning content. Currently, game type is something selected by the game designer or programmer. Previous research study has mentioned that game type and lesson content is a critical area that requires more research. At the moment, a teacher or game designer is not too sure what game type would be best to teach a lesson on Geography or Mathematics. In fact, it has been found in the previous study that game type does have a significant impact on learning outcome and experience. To capitalize on the research gap for game type, we have therefore embarked on Stage 2 of our study to use artificial intelligence (AI). A machine learning model is employed to predict the evaluation score of the game type of mobile educational game employed to teach a subject lesson and to recommend the best game type for teaching the lesson. We then proceeded to Stage 3 and evaluated the performance of the AI model by creating a test set of twenty games and twenty undergraduates were recruited at an Indonesian university to evaluate the games. The average score of all mobile games evaluations is above the average of 3.5, thus proving the hypothesis H1 set out for Stage 3.

Actions (Repository Staff Only)

Item Control Page