Methodological approaches of machine learning and adaptive personalization in educational gaming platforms

A case study of gamified EdTech systems for personalized learning

Abstract

This article examines methodological approaches in machine learning (ML), educational data mining (EDM), and adaptive gamification on digital educational platforms, based on a systematic review of 36 peer-reviewed studies published between 2020 and 2025. The study identified the most frequently applied ML models, personalized learning approaches, and game mechanics that enhance student motivation and learning outcomes. The literature search was conducted in the Scopus, IEEE Xplore, SpringerLink, MDPI, ScienceDirect, and ResearchGate databases, following strict inclusion criteria focused on the concrete implementation of technologies in the educational process. The analysis revealed the active use of graph neural networks, reinforcement learning, gradient boosting, and deep learning architectures for adaptive learning and behavioral modeling. At the same time, it was found that adaptive gamification, when combined with real-time analytics, significantly improves long-term student engagement. As a result of the comparative analysis of ML methodological approaches and adaptive personalization, a model of interaction between the player, the EdTech game, and an RL-based personalization system was proposed.