Towards an Intelligent Model for Evaluating Serious Games

Abstract

Serious games are effective educational tools used in higher education to provide practical learning opportunities to students. However, few research works have focused on evaluating serious games as a project for developing a tool dedicated to use in a formative context. This document proposes an intelligent evaluation model that not only allows for the evaluation of serious games but also facilitates their integration into teaching practice. The model is designed around four dimensions, and their measurement criteria are well defined. Fuzzy decision-making methods were used to weight the criteria, and supervised machine-learning algorithms were considered to minimize the evaluator’s bias. The proposed model provides a more objective and consistent solution for evaluating serious games, reducing the impact of evaluators’ biases and subjective preferences on the weightings of the different evaluation dimensions. The multi-output support vector regression (M-SVR) model can be used flexibly and adapted to different contexts and applications, offering a more effective and reliable solution for evaluating serious games.