Scalable AI-driven Framework for Comprehensive Gameplay Characterization

GAMECHAR addresses a critical gap in gameplay characterization of video games by developing scalable, data-driven methods that link gameplay mechanics to player experiences. Current methodologies face limitations due to fragmented approaches, small datasets, and challenges in comparing human-generated and AI-generated gameplay data. Existing tools struggle to standardize the characterization of gameplay, making it difficult to extract meaningful insights from diverse games. Additionally, the emergent and interactive nature of gameplay requires systematic, scalable solutions for comprehensive analysis.To overcome these challenges, GAMECHAR will develop innovative methodologies to characterize gameplay data, establishing standardized descriptors and classification models. This approach will automate gameplay data collection, processing, and feature engineering, enabling researchers and developers to distinguish between restorative and harmful gameplay. The project will contribute to gameplay evaluation and advance the quality of gameplay developed to address societal needs, e.g., healthcare. The project contributes to the EU goals for AI excellence and digital health. Key outcomes will include a comprehensive gameplay data repository, AI agents capable of automated gameplay testing, and classification models that generalize across different game environments. Methods and processes will provide researchers with tools for gameplay analysis, enabling more effective comparative studies and predictors of player experience. The PI’s pathways to regulatory authorities have the potential for broad societal impact, applying the developed approach in policy and regulation of exploitative gameplay.