Gamification with Natural Language Processing for Educational Videogame Modeling

Traditional education has been based for many years on monotonous teaching techniques and practices to assess students' skills and knowledge in the classroom, which interferes with the evolution and adaptation of education using recent technologies. To create new learning methods using intelligent and digital tools, this paper presents an architecture that uses gamification and natural language processing techniques as an alternative to propose new learning and feedback methods for high school students. The main objective is to motivate students to obtain new knowledge or study what they have already learned in the classroom while playing video games, and for teachers to obtain relevant data on the learning level of each student. As a result of the proposed architecture, a platform video game for mobile devices has been developed, which performs quizzes on specific topics of biology to high school level education. Users give answers through voice commands, which are processed with the VOSK library, and then compared with other options using the cosine similarity method. Finally, an algorithm leverages the scores acquired from video game actions on scenes and questionnaires to synergistically derive a comprehensive metric that assesses the user's proficiency in learning. This methodology was validated according to several experiments, and in future work, it will be applied to control and test groups of students to validate the level of learning obtained using the proposed platform.