Ontology Enrichment of Video Games with LLMs
Oguzhan Menemencioglu, Ahmet Selim Kucukkara, D Chaves-Fraga, None Heibi, D Garijo, D Collarana, A Salatino, S Vahdati
proceeding (Uluslararası, Hakemli) | Yazar(lar): Oguzhan Menemencioglu, Ahmet Selim Kucukkara, D Chaves-Fraga, None Heibi, D Garijo, D Collarana, A Salatino, S Vahdati | Yıl: 2025 | Dergi/Kurum: SEMANTICS-PDWT 2025 | İndeks: WOS.ISTP | Anahtar Kelimeler: Ontology, video games, video game universes, ontology enrichment,
Video games have evolved into a prominent cultural and economic phenomenon, increasingly recognized as a powerful medium across diverse domains, including art, education, and social interaction. This evolution has necessitated a more nuanced and systematic analysis of game universes, along with the development of methodological frameworks to facilitate their understanding. Ontology, defined as a formal representation of concepts and the relationships among them within a given domain-whether for human interpretation or machine processing-has been employed as a means to model and describe the structural and semantic components of video games. The primary objective of this study is to identify and articulate existing gaps in the literature like relations, roles, locations, tasks, and voice acting concerning the ontological modeling of video game universes and to address these gaps by proposing an ontologically grounded perspective for analyzing virtual worlds. This paper introduces ontological models for core game elements, including character relationships, role typologies, spatial representations, mission structures, and voice acting. Each proposed ontology is examined in terms of its capacity to enhance our understanding of the underlying dynamics of game worlds. Furthermore, the study explores the integration of Large Language Models (LLMs) as a means of enriching ontological structures, with an emphasis on moving beyond manual approaches to support scalability and semantic depth.