This paper presents a structural topic modeling analysis of the computational design literature spanning from 2000 to 2024. By analyzing 12,550 scientific publications from diverse disciplines, this research maps the thematic structure, temporal evolution, interdisciplinary knowledge migration, and geographical distribution of research in computational design. The findings reveal 15 distinct topics that characterize the field, with significant shifts over time from architectural generative design and optimization towards artificial intelligence-based approaches and creative coding. Notable patterns of knowledge migration between disciplines are identified, particularly in the adoption of AI techniques from computer science into design, and digital fabrication methods from engineering into architecture. The analysis also highlights geographical specializations, with Nordic countries focusing on sound design and user experience, East Asian countries on visual design and interactive narrative, and North America on AI-based approaches. This research reveals that the rise of artificial intelligence-based approaches is fundamentally altering not only the technological toolset of the field but also the very nature of interdisciplinary knowledge flow.
As the digital transformation driven by the Fourth Industrial Revolution fundamentally reshapes creative
practices, computational design stands at the center of this change. In this context, understanding how
the knowledge and methods of diferent disciplines interact and transform one another is of critical
importance for mapping the future trajectory of the field.The integration of computational methods
into design processes has transformed creative practices across diferent disciplines. From architecture
to product design, visual arts to music, computational design approaches have enabled new forms of
expression, expanded the boundaries of what is creatively possible, and facilitated innovative workflows
that combine human creativity with algorithmic processes. This intersection of design, computation,
and creativity has given rise to a rich interdisciplinary landscape that continues to evolve rapidly with
technological advancements [
Computational design serves as a bridge, connecting knowledge and practices across disciplinary
boundaries. Architects utilize generative algorithms to shape physical spaces [
Recent advances in artificial intelligence and machine learning have opened new horizons in
computational design and creativity. Generative adversarial networks (GANs), transformer models, and other
deep learning techniques have expanded the toolset of designers and creators, enabling new forms of
human-machine creative collaboration [
However, this rapid integration of AI into creative processes has raised fundamental questions about the future of computational design. Concerns about job displacement in creative industries, debates over authorship and authenticity in AI-assisted work, and the need to preserve human creative agency while leveraging computational capabilities have become central to contemporary discourse in the field.
Despite the growing literature on computational design, systematic analyses of how knowledge
and methodologies transfer across diferent disciplinary domains remain limited. Goel and Davies [
To define the research scope, this study examined review articles published in reputable journals and
conferences (ICCC, EvoMUSART, ACM Creativity&Cognition, SIGGRAPH, CAAD Futures, etc.) over
the past fifteen years. Based on this systematic approach, the following methodological framework
was implemented to ensure comprehensive coverage and analytical rigor. Following Kitchenham and
Charters [
Structural Topic Modeling, developed by Roberts et al. [23], was used for text analysis. Unlike traditional topic modeling methods, STM allows document characteristics (covariates) to influence both topic content and topic prevalence. Through diagnostic measures including semantic coherence and exclusivity metrics, K=15 topics was determined as the optimal number providing balance between interpretability and granularity [24], [25].
Three main covariates were incorporated: temporal (publication year, 2000-2024), disciplinary (field categorization), and geographical (country of first author afiliation) [ non-linear (spline) function of year [29].
For analyzing interdisciplinary knowledge transfers, the analysis identified the dominant topic for each document, calculated topic distributions for each discipline-year slice, and modeled transitions between consecutive time slices as a weighted directed network [30], [31], [32]. This approach allowed visualization of topic migrations between disciplines over time.
Geographical distribution analysis utilized scientific mapping methodology [ 33], calculating countrybased topic distributions. Temporal analyses examined changes in topic probabilities over time in annual and 2-year slices [34]. All analyses and visualizations were performed using R programming language with specialized packages [35], [36], [37]. All analyses used R programming [35], [36], [37]. . Limitations include English-only publications and Scopus coverage, potentially underrepresenting Global South research and non-Western academic traditions
3.0.1. Thematic Structure of Computational Design The Structural Topic Modeling (STM) analysis identified 15 fundamental topics within the computational design literature. These topics and their most distinctive terms are presented in Table 1. The analysis revealed that the topics with the highest prevalence are Architectural Generative Design (T1) (14.3%), AI-Based Design (T3) (11.7%), Manufacturing and Fabrication (T4) (9.8%), and Creative Coding (T2) (9.1%). 15 topics identified by STM analysis and their most distinctive terms.
Topic ID Label Most Distinctive Terms T1 Architectural Generative Design parametric, geometric, algorithmic, façade, architect, form, optimization, script, pattern, generative T2 Creative Coding code, program, creative, software, interact, artist, tool, platform, interface, develop T3 AI-Based Design ai, learn, generate, neural, gan, deep, network, intelligent, model, train T4 Manufacturing and Fabrication print, manufacturing, fabrication, 3d, material, additive, product, structure, layer, robot T5 Visual Design and Typography graphic, typography, visual, layout, font, composition, create, image, element, text T6 Sound and Music Production sound, music, audio, composition, instrument, generate, acoustic, perform, signal, frequency T7 User Experience in Design user, experience, interact, interface, product, evaluate, design, human, usability, test T8 Knowledge-Based Systems ontology, semantic, knowledge, inference, rule, database, represent, reason, schema, query T9 Optimization and Simulation optimization, simulation, algorithm, performance, function, solution, constraint, objective, eficiency, search T10 Computational Creativity Theory creativity, cognitive, computational, process, human, theory, conceptual, automation, intent, system T11 Design Tools and Workflows workflow, tool, bim, process, collaboration, integration, platform, management, project, implementation T12 Computational Design in Education education, pedagogy, learn, teach, student, curriculum, skill, course, academic, workshop T13 Data-Driven Design data, visual, analysis, model, structure, information, extract, pattern, cluster, set T14 Urban and Spatial Design urban, spatial, city, plan, space, map, environment, model, analysis, geographic T15 Games and Interactive Narrative game, play, narrative, interact, virtual, character, scene, environment, player, story
The analysis revealed significant temporal shifts in topic prevalence between 2000 and 2024 ( Figure 2). The most pronounced changes manifested as a rise in the popularity of some topics and a decline in others. Notably, AI-Based Design (T3) exhibited the most remarkable growth, increasing its share from 5.2% in the early 2000s to 18.7%. Similarly, Creative Coding (T2) gained prominence, rising from 4.9% to 12.3%. In contrast to this growth, the prevalence of Architectural Generative Design (T1), which was dominant in the field’s early years, decreased from 18.5% to 11.4%. During the same period, Optimization and Simulation (T9) also lost its earlier popularity, falling from 12.1% to 7.3%. On the other hand, topics such as Sound and Music Production (T6) and Computational Design in Education (T12) maintained relatively stable shares throughout the analyzed period. The introduction of artificial intelligence and deep learning techniques into the design field, especially post-2017, has significantly afected this topic distribution. The rise of generative models (GANs and difusion models) explains the dramatic increase in the prevalence of AI-Based Design.
The examination of interdisciplinary topic migrations reveals prominent patterns that highlight the direction and nature of knowledge flow within the field (see Figure 3). One of the most significant transitions, becoming particularly evident after 2015, is the transfer of the AI-Based Design (T3) topic from Computer Science to Design, which demonstrates how the design discipline has adopted and adapted artificial intelligence techniques. A similar dynamic was observed in the continuous flow of the Manufacturing and Fabrication (T4) topic from Engineering to Architecture, reflecting the integration of computational design with physical production processes. Other important migrations include the adaptation of Creative Coding (T2) from Design to Arts & Humanities and the way problems in Architectural Generative Design (T1) from Architecture have opened new research areas for Computer Science. These migration patterns reflect the dynamic nature of interdisciplinary interaction in the computational design field, with a bidirectional flow of knowledge observed, particularly in the topics of AI-Based Design and Creative Coding.
The geographical analysis revealed distinctive regional research focuses, as shown in Table 2. This geographical distribution reflects unique regional research traditions and priorities. For instance, Nordic countries (Sweden, Finland, Denmark) stand out in Sound and Music Production (T6) and User Experience in Design (T7). East Asian countries (Japan, South Korea) exhibit a strong focus on Visual Design and Typography (T5) and Games and Interactive Narrative (T15). In contrast, North America (USA, Canada) dominates in technology-intensive topics such as AI-Based Design (T3), Optimization and Simulation (T9), and Data-Driven Design (T13).
The identification of 15 distinct topics reveals computational design as a field encompassing both theoretical exploration and practical application across multiple domains. The temporal analysis demonstrates that architectural generative design dominated the field’s early development, establishing computational methods that subsequently influenced other disciplines. This finding validates Woodbury’s [ 38] observations about parametric design’s transformative efect while documenting its role as a foundational element in the field’s evolution. This foundational stability underwent significant change in the field’s later development.
The results also support Menges and Ahlquist’s [39] emphasis on the integration of computational design with material systems and fabrication processes. The high prevalence of T4 (Manufacturing and Fabrication) indicates a close relationship between digital design and physical production.
The substantial growth of AI-Based Design represents the most significant transformation
documented in this analysis, accelerating particularly after 2017 with mainstream AI tool availability. This
shift suggests reorientation toward human-machine collaboration while raising critical questions about
creative agency and professional displacement. The trend validates McCormack et al.’s [
The documented migration patterns reveal computational design as an active knowledge mediator
rather than isolated practice. Three pathways emerged: AI techniques flowing from Computer Science
to Design, manufacturing knowledge migrating from Engineering to Architecture, and Creative Coding
establishing bidirectional Design-Arts exchange. These patterns challenge Goel and Davies’s [
The flow of Manufacturing and Fabrication from Engineering to Architecture supports Kolarevic and Klinger’s [40] thesis on the reintegration of architectural design and production processes, linked to the adoption of digital fabrication technologies in architectural practice.
While the rise of AI-based approaches is the most striking trend, other important developments were observed. The increase in Data-Driven Design reflects the growing use of big data in design processes, as emphasized by Ofenhuber and Ratti [ 41], indicating evolution toward evidence-based design approaches.
The relationship between User Experience in Design and Games and Interactive Narrative reflects the "gamification" trend identified by Deterding et al. [ 42], developing predominantly under the leadership of Nordic research groups.
The stable presence of Computational Design in Education confirms Oxman’s [
These converging trends suggest three strategic directions for computational design’s evolution. First, AI-driven transformation requires frameworks preserving human creative agency while leveraging computational capabilities. Second, successful migration patterns provide templates for fostering exchange between isolated domains. Third, addressing geographical participation gaps necessitates systematic support for underrepresented regions. These findings indicate that computational design’s advancement depends on thoughtful navigation of technological, collaborative, and cultural dimensions
Future research could extend this work with full-text analyses, citation network analyses, and examination of non-textual content (images, codes, models). Deeper analysis of interdisciplinary topic migrations could provide insights into the mechanisms driving these knowledge transfers.
This study maps the thematic structure, temporal evolution, interdisciplinary knowledge migrations, and geographical distribution of the computational design field through Structural Topic Modeling analysis of 12,550 scientific publications. The findings show that computational design is a versatile and dynamic interdisciplinary field spanning architecture, computer science, art, and engineering.
This study provides valuable insights for computational design educators, researchers, and practitioners by identifying opportunities for interdisciplinary collaboration, emerging trends, and future research directions. It contributes to a deeper understanding of how computational design has transformed creative applications across diferent disciplines and how it continues to evolve as a field at the intersection of technology, creativity, and design.
This study acknowledges several limitations. The analysis focuses on English-language publications in Scopus, potentially underrepresenting non-Western research traditions. Additionally, geographical categorization based on first author afiliation may not fully capture international collaborations, and the reliance on titles and abstracts rather than full-text content may limit thematic detail.
Ultimately, this study demonstrates that computational design has evolved beyond a collection of tools and techniques to become a dynamic ecosystem that fundamentally challenges traditional disciplinary boundaries. The documented shift from architectural parametricism to AI-driven creativity, coupled with systematic knowledge migration patterns, reveals a field in profound transformation. As artificial intelligence continues to reshape creative practices, understanding these interdisciplinary dynamics becomes essential for educators developing future curricula, researchers identifying promising directions, and practitioners navigating an increasingly complex technological landscape. Computational design’s future lies not in any single discipline but in the spaces between them—where technology, creativity, and human insight converge to expand the boundaries of what is possible.
In this study, the author used ChatGPT-4o3 for translation, language checking, and coding processes. All content was subsequently reviewed and edited by the author, who takes full responsibility for the publication. [19] H. Arksey, L. O’Malley, Scoping studies: Towards a methodological framework, International
Journal of Social Research Methodology 8 (2005) 19–32. [20] D. M. Blei, Probabilistic topic models, Communications of the ACM 55 (2012) 77–84. [21] M. E. Roberts, B. M. Stewart, D. Tingley, stm: An r package for structural topic models, Journal of
Statistical Software 91 (2019) 1–40. [22] H. M. Wallach, I. Murray, R. Salakhutdinov, D. Mimno, Evaluation methods for topic models, in: Proceedings of the 26th Annual International Conference on Machine Learning, 2009, pp. 1105–1112. [23] M. E. Roberts, B. M. Stewart, D. Tingley, C. Lucas, J. Leder-Luis, S. K. Gadarian, B. Albertson, D. G.
Rand, Structural topic models for open-ended survey responses, American Journal of Political Science 58 (2014) 1064–1082. [24] D. M. Blei, A. Y. Ng, M. I. Jordan, Latent dirichlet allocation, Journal of Machine Learning Research 3 (2003) 993–1022. [25] D. Mimno, H. Lee, Low-dimensional embeddings for interpretable anchor-based topic inference, in: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, 2014, pp. 1319–1328. [26] D. Binkley, D. Heinz, D. Lawrie, J. Overfelt, Understanding lda in source code analysis, in:
Proceedings of the 22nd International Conference on Program Comprehension, 2014, pp. 26–36. [27] N. Hara, P. Solomon, S.-L. Kim, D. H. Sonnenwald, An emerging view of scientific collaboration: Scientists’ perspectives on collaboration and factors that impact collaboration, Journal of the American Society for Information Science and Technology 54 (2003) 952–965. [28] K. Borner, R. Klavans, M. Patek, A. M. Zoss, J. R. Biberstine, R. P. Light, V. Lariviere, K. W. Boyack,
Design and update of a classification system: The ucsd map of science, PloS One 7 (2012) e39464. [29] L. Keele, Semiparametric Regression for the Social Sciences, John Wiley & Sons, Hoboken, NJ, 2008. [30] J. Han, F. Shi, L. Chen, P. R. Childs, A computational tool for creative idea generation based on analogical reasoning and ontology, Artificial Intelligence for Engineering Design, Analysis and Manufacturing 32 (2018) 462–477. [31] E. Yan, Y. Ding, Scholarly network similarities: How bibliographic coupling networks, citation networks, cocitation networks, topical networks, coauthorship networks, and coword networks relate to each other, Journal of the American Society for Information Science and Technology 63 (2012) 1313–1326. [32] V. D. Blondel, J.-L. Guillaume, R. Lambiotte, E. Lefebvre, Fast unfolding of communities in large networks, Journal of Statistical Mechanics: Theory and Experiment 2008 (2008) P10008. [33] L. Leydesdorf, O. Persson, Mapping the geography of science: Distribution patterns and networks of relations among cities and institutes, Journal of the American Society for Information Science and Technology 61 (2010) 1622–1634. [34] T. L. Grifiths, M. Steyvers, Finding scientific topics, Proceedings of the National Academy of
Sciences 101 (2004) 5228–5235. [35] R. C. Team, R: A language and environment for statistical computing, Technical Report, R Foundation for Statistical Computing, Vienna, Austria, 2021. [36] G. Csardi, T. Nepusz, The igraph software package for complex network research, InterJournal,
Complex Systems 1695 (2006) 1–9. [37] H. Wickham, ggplot2: Elegant Graphics for Data Analysis, Springer-Verlag, New York, 2016. [38] R. Woodbury, Elements of Parametric Design, Routledge, London, 2010. [39] A. Menges, S. Ahlquist (Eds.), Computational Design Thinking: Computation Design Thinking,
John Wiley & Sons, Chichester, 2011. [40] B. Kolarevic, K. R. Klinger (Eds.), Manufacturing Material Efects: Rethinking Design and Making in Architecture, Routledge, New York, 2008. [41] D. Ofenhuber, C. Ratti, Decoding the City: Urbanism in the Age of Big Data, Birkhäuser, Basel, 2014. [42] S. Deterding, D. Dixon, R. Khaled, L. Nacke, From game design elements to gamefulness: Defining ’gamification’, in: Proceedings of the 15th International Academic MindTrek Conference: Envisioning Future Media Environments, 2011, pp. 9–15.