Big data and AI for decision-making purposes in the metaverse: A cross-domain review of architectures, applications, and governance
Multidisciplinary Reviews · 09.10.2026 · 10 (4) · 2027229
DOI: 10.31893/multirev.2027229 ↗Kurz gefasst
Fachartikel in der Zeitschrift Multidisciplinary Reviews, online erschienen am 09.10.2026. Verfasst von Leonidas Theodorakopoulos und 2 weiteren Autor:innen. Die Arbeit nennt 153 Literaturangaben.
Abstract (Originalsprache)
This review examines how big data and artificial intelligence are being mobilized to support decision-making in metaverse and metaverse-adjacent environments. Rather than treating the metaverse as an interface novelty, the paper frames it as a decision environment in which persistent interaction, embodied participation, and multimodal sensing tighten the loop between observation and intervention. Using a transparent narrative review approach, the synthesis applies a consistent decision-making taxonomy across nine domains (industrial and manufacturing operations, smart cities, healthcare, retail and commerce, supply chain and logistics, energy and utilities, finance, education, tourism). For each domain, evidence is organized around decision level, decision mode, time requirements, data context, and action type, enabling comparability without collapsing domain-specific consequences into a single maturity scale. Across domains, the most salient pattern is that decision value depends less on immersion itself and more on the integration of digital representations (often via digital twins), AI inference, and bounded action pathways. The review highlights metaverse-specific constraints that shape decision reliability: measurement validity challenges in multimodal signals, proxy gaming when engagement metrics substitute for decision impact, identity and consent boundaries that constrain permissible inference, and latency/edge limits that compress feasible action sets. Building on these observations, the paper proposes a reference architecture organized as a data–AI–simulation–action loop with explicit fast and slow paths, emphasizing safe-state behavior, fallback logic, and escalation rules as architectural requirements for real-time decisions. A compact cross-domain synthesis consolidates recurring gaps as testable absences in current work, including unclear decision objects, weak semantic and identity alignment, limited uncertainty communication, under-specified constraint handling, insufficient audit trails, and inadequate robustness to drift and adversarial manipulation. The paper concludes with governance and ethics considerations anchored in established regulatory and standards frameworks, and outlines research directions aimed at decision-centered evaluation protocols, contestable immersive interfaces, and governance-ready metaverse decision systems.
Zitieren
APA
Theodorakopoulos, L., Theodoropoulou, A., & Bakalis, A. (2026). Big data and AI for decision-making purposes in the metaverse: A cross-domain review of architectures, applications, and governance. Multidisciplinary Reviews, 10(4), 2027229. https://doi.org/10.31893/multirev.2027229
Vancouver
Theodorakopoulos L, Theodoropoulou A, Bakalis A. Big data and AI for decision-making purposes in the metaverse: A cross-domain review of architectures, applications, and governance. Multidisciplinary Reviews. 2026;10(4):2027229. doi:10.31893/multirev.2027229
BibTeX
@article{Theodorakopoulos2026,
author = {Theodorakopoulos, Leonidas and Theodoropoulou, Alexandra and Bakalis, Aristeidis},
title = {Big data and AI for decision-making purposes in the metaverse: A cross-domain review of architectures, applications, and governance},
journal = {Multidisciplinary Reviews},
year = {2026},
volume = {10},
number = {4},
pages = {2027229},
doi = {10.31893/multirev.2027229}
}
EUVIMED