Integration of Generative Adversarial Networks into Participatory Architectural Practice: A Scoping Review
12th International Conference on New Trends in Architectural and Interior Design (12th ICNTAD’26), Barcelona, İspanya, 10 - 12 Nisan 2026, ss.167-176, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Basıldığı Şehir: Barcelona
- Basıldığı Ülke: İspanya
- Sayfa Sayıları: ss.167-176
- Çukurova Üniversitesi Adresli: Evet
Özet
This scoping review systematically examines the integration of Generative Adversarial Networks (GANs) into participatory architectural practice through a three-stage bibliometric analysis. Using the Web of Science database and VOSviewer software, the study analyzed 290 publications across three progressively focused searches: (1) artificial intelligence and architectural design (n=44), (2) participatory design and artificial intelligence (n=146), and (3) GANs and architectural design (n=100). The analysis reveals a critical bifurcation in contemporary research. Stage 1 demonstrated that AI-architecture literature centers on generative tools (Midjourney, Stable Diffusion, GANs) and technical methodologies but completely lacks participatory or user-centered approaches. Stage 2 established a robust participatory design and AI literature emphasizing human-centered frameworks, conversational interfaces, and ethical considerations (algorithmic justice, responsible AI), yet this body of work exhibits zero engagement with architectural or spatial design applications. Stage 3 confirmed that GANs-architecture literature is technically sophisticated but entirely devoid of participatory design concepts, user engagement frameworks, or collaborative methodologies. Keyword co-occurrence network analyses revealed disciplinary isolation: technical AI-architecture research proceeds independently from human-centered participatory AI research, with minimal cross-pollination. Temporal analysis showed convergent trends, with significant publication increases from 2022 and peaks in 2025 for both AI-architecture and GANs-architecture domains, yet the persistent mutual absence of participatory concepts in architectural AI literature and architectural contexts in participatory AI literature indicates systematic neglect. The findings definitively validate a critical research gap: despite growing interest in both AI-assisted architectural design and participatory AI systems, the integration of GANs into participatory architectural practice remains unexplored territory. This review establishes the conceptual foundations for reconceptualizing GANs not merely as technical tools but as collaborative design actors in human-centered architectural processes. The emerging research agenda calls for interdisciplinary synthesis bringing together technical AI capabilities, participatory design methodologies, and architectural domain expertise to enable more inclusive, responsive, and democratized AI-assisted architectural practice.