This study comprehensively evaluates the contributions of text-to-image artificial intelligence (AI) systems to interior architectural design using multi-criteria decision-making (MCDM) methods. Six platforms, DreamStudio, MidJourney, Leonardo AI, Artbreeder, Craiyon and DALL-E, were examined by an expert panel of interior designers and architects and analyzed using MCDM techniques including TOPSIS, AHP, VIKOR, ELECTRE, and PROMETHEE. The analyses revealed DreamStudio’s strengths in spatial organization and material harmony, while MidJourney stood out for its ability to generate dynamic, balanced, and visually diverse compositions. Correlation analysis among the MCDM methods enhanced the reliability of the findings, particularly highlighting a strong alignment between TOPSIS and AHP (r=0.997). The study demonstrates the strong aesthetic potential of current AI systems but underscores their limitations in fundamental design elements like spatial logic and cultural relevance. Aesthetic biases and ethical considerations are also addressed. Future research should integrate user experience and designer perspectives to explore a more meaningful and holistic integration of AI into interior design processes.

Energy efciency in buildings has become an increasingly important design criterion in the context of sustainable architecture. In this context, building envelope components such as window systems must be thoroughly evaluated in terms of energy performance, initial costs, and long-term total cost. This study draws attention to the limitations of the commonly adopted one-dimensional approaches in selecting window glazing systems and offers a comprehensive analysis across three key parameters: Investment cost, energy consumption, and life cycle cost. A total of 64 scenarios are developed for a theoretical residential building located in Istanbul. Comparisons are made across three key performance criteria, taking into account different facade orientations (north/south) and window-to-wall ratios (30%/60%). The results reveal that the optimal window systems differ across the performance criteria, suggesting that selecting a system based solely on energy consumption or investment cost may result in suboptimal or misleading outcomes. Systems with low energy consumption often come with higher investment costs, which may limit the potential to minimize overall life cycle cost. Double-glazed systems with solar low-e coatings tend to yield more favorable life cycle cost
outcomes, thanks to their relatively low investment costs and balanced overall performance. The study underscores the importance of adopting a multidimensional approach in window system selection and offers practical guidance to decision-makers aiming for cost-effective design strategies.

The aesthetic evaluation of architectural computer renderings has traditionally remained subjective and dependent on personal, situational, and cultural factors. Within this research, we investigate if deep learning (DL) can be utilized to provide a scientific data-driven solution for approximating the perceived aesthetics in architecture. Our focus is on standalone house designs and uses a dataset of 1,438 computer-rendered competition entries off the Arcbazar website, assigned a rating by professional architects for visual quality. In this research, "aesthetic evaluation" refers to the numerical scores given to the attractiveness to architectural renderings. Our dataset of renderings was standardized through image preprocessing and paired with averaged expert scores. A supervised convolutional neural network (CNN) regression model was then trained and validated using three-fold cross-validation. Model accuracy was established using standard measures of regression (MAE, MSE, RMSE, and R²). Results indicate that the model was able to predict aesthetic scores with high validity. While the findings demonstrate the validity of DL models to evaluate architectural renderings, the following limitations should be pointed out: The dependence on rendered views, assessment of just one building type, and expertise based on raters from one platform. Future research will have to expand on the aspects of differing building types, cultural contexts, and multimodal inputs. The incorporation of explainable AI methods will further assist in identifying which visual features contribute most to aesthetic prediction. This work establishes a proof-of-concept framework for integrating deep learning into architectural evaluation, supporting an extensible system that allows for design competition and decision-making. Apart from predictive scoring, such models are well-suited to be integrated with generative design frameworks that will enable the generation of novel architectural proposals optimized in aesthetic quality.

Understanding the dynamics of the real estate market has emerged as a pivotal concern in urban economics for ensuring sustainable land management and effective investment strategies. The spatial heterogeneity of housing market determinants gives rise to variations in market activity and significant spatial differences in property values. However, global regression models are constrained in their ability to capture this heterogeneity and the spatial autocorrelation of housing prices. The aim of this study is to identify the spatial variability of housing prices and the potential factors influencing them. To this end, the study employs the Geographically Weighted Regression (GWR) model, which enables the analysis of spatial heterogeneity. In addition to conventional structural variables (floor area, age, heating type, number of floors, and floor level), measurable indicators, including network-based accessibility metrics (connectivity, betweenness, and closeness), distance to the central business district, and the remotely sensed Normalized Difference Vegetation Index (NDVI), are integrated into the model. The findings reveal the complexity of the housing market in Erzurum, showing that newly developing peripheral areas form high-priced clusters that reshape the center-periphery dynamics. While structural variables, including floor area and building age, emerge as dominant factors across the city, environmental determinants vary considerably by location. It is noteworthy that network-based accessibility metrics are critical infrastructural variables that shape market heterogeneity. NDVI highlights the decisive role of vegetation density and accessible and functional urban green spaces in determining housing values. In conclusion, this study offers novel insights into the role of environmental and infrastructural metrics in real estate research and provides guidance for policymakers in regulating housing values and designing more sustainable urban planning strategies.

The concept of studentification, often discussed as a form of gentrification, has been predominantly studied in the United Kingdom, the United States, and Canada, where globally prominent universities and market-driven higher education systems prevail. While recent studies have begun to introduce geographical diversity, the planning dimensions of studentification remain underexplored. This paper examines the dynamics of studentification in the case of İzmir Katip Çelebi University in İzmir, Türkiye, highlighting differences between the Anglophone literature (UK, US, Canada) and the Turkish contexts. In Türkiye, rapid university expansion, limited state-provided accommodation, and a planning system that has yet to integrate student housing into its strategic framework have shaped distinct conditions for studentification. The findings indicate that local authorities have not developed proactive, plan-led responses, and strategic and spatial plans lack measures to address the social and spatial consequences of the growing student population. Drawing insights from the United Kingdom, the United States, and Canada, this paper proposes recommendations to enhance regulatory frameworks, strengthen compliance standards, and support municipalities in addressing the challenges of studentification in Türkiye.

Urban squares are crucial for urban life, enhancing city attractiveness and livability; nevertheless, many in Istanbul lack pedestrian-oriented design and sufficient infrastructure. Relatedly, several projects have been implemented in urban squares to overcome these issues. However, designing urban squares leads to both improvements and controversies, reshaping not only the physical environment but also how individuals engage with public space, and understanding pre/post dynamics of interventions provides comprehensive feedback and helps municipalities to resolve further needs. This study explores the impact of spatial changes on pedestrian behavior in Besiktas Barbaros Square (Istanbul/Turkey) and its surroundings. With a three-stage methodology, this research first identifies physical transformations by analyzing satellite imagery, street views, and field studies. Second, pedestrian flows and stationary activities are observed via manual video recordings (10 minutes, weekday/weekend, 2022-2024), capturing the site both before (once the steel overpass was removed and before the site was redeveloped) and after interventions. These observations are mapped and analyzed using QGIS. Third, perspectives from professionals (n=31) in spatial fields -urban planners, architects, and landscape architects- are gathered through surveys and open-ended responses. The findings focused on the eastern part of the square due to observational constraints, which reveal that adding urban furniture impacts dynamic/stationary activities and enhanced social interactions. The new eagle sculpture also emerged as a focal point. According to professionals’ evaluations, interventions contributed to partial improvements in several aspects between pedestrians and the space. However, considering location-based potentials, green space use and the spatial connection with the urban coastline still need to be improved.

This study explores the potential of text-to-3D model AIs in designing urban furniture, with a focus on bus stops, street lamps, and benches, and provides a comparative evaluation of four prominent AI tools: Luma Genie, Meshy, Tripo, and Deepai. A diverse poll of architects, urban planners, industrial designers, and students assessed the outputs based on key criteria: Aesthetic appeal, texture detail, form detail, and technical consistency and feasibility. The comparative analysis revealed that Meshy consistently outperformed the other platforms across all criteria, achieving the highest overall score of 4.09. Meshy's success is attributed to its high performance in visual creativity, structural sophistication, and spatial awareness. Conversely, Deepai lagged significantly, notably lacking in functional logic, spatial awareness, and technical consistency, resulting in the lowest overall score of 1.69. While Luma Genie and Tripo showed balanced performance, they did not match Meshy's degree of structural and aesthetic intricacy. This study highlights the current limitations of text-to-3D AI, emphasizing that platform-specific features like customization and technical control play a critical role in generating feasible architectural outputs for the future of urban design.

This study analyzes the impact of digitalization and interdisciplinary design approaches on production processes in architectural offices in the 21st century, exploring the implications of this transformation on institutional structure, actor representation and collaboration models, particularly in relation to theory and practice. Architectural offices are considered dynamic spaces where digital technologies, beyond being mere technical tools are also positioned as structural components that transform production culture, while interdisciplinarity is inherently organized within operational practices. The research is based on the content analysis of 45 articles published between 2005 and 2025 and on the comparative evaluation of BIG, MVRDV and Herzog & de Meuron from Europe and Tabanlıoğlu, GAD, and Erginoğlu & Çalışlar from Türkiye. In the content analysis, themes were systematically coded through frequency and co-occurrence relationships, resulting in a conceptual structure concentrated around participatory design, interdisciplinary teamwork and the ecology framework. This structure made visible how production tools such as BIM integration, digital twins, AI-based workflows and parametric modeling are positioned across both computational and organizational dimensions. The findings reveal that in European offices, digital tools are integrated into institutionalized computational cycles from the early stages of design, whereas in the Türkiye examples, these tools diversify more flexibly in relation to local context, multi- actor negotiation and project-based strategies. The study demonstrates that the digitalizing production culture reshapes design processes from linear sequences into data-driven, multilayered and iterative structures. This framework suggests that, for the future of architectural practice, data-driven participation models and the integration of computational–interdisciplinary processes into the early stages of design are becoming increasingly central design approaches.