Analysis of the Historically Compatibility of AI-Assisted Urban Furniture Design Using the Semantic Differentiation Method: The Case of Elazığ Harput

Abstract

This study examined the historical compatibility of urban furniture in Harput Sarahatun Mosque Square, Elazığ, Türkiye. It evaluated AI-generated modern and classical-style alternatives using the Semantic Differentiation Method. The research aimed to compare existing furniture with AI-assisted designs and identify key attributes influencing historical and spatial integration. The methodology consists of four stages: (1) defining adjective pairs to assess historical compatibility through expert opinions and literature review; (2) photographing existing urban furniture and generating AI-assisted modern and classical-style urban furniture (benches, trash cans, and lighting elements); (3) determination expert opinion using the survey; (4) statistical analysis of results through descriptive statistics and explanatory factor analysis (EFA). The study, which was conducted online in February 2025, involved 31 experts from the architecture and landscape architecture disciplines. The findings show that existing furniture is mainly judged by practicality and usability, with limited attention to historical integration. Modern AI-generated designs emphasize innovation, minimalism, and contemporary aesthetics. In contrast, classical-style AI-generated furniture is appreciated for its historical compatibility, cultural resonance, and aesthetic harmony. Experts favored classical alternatives for their alignment with traditional urban character. The results highlight the need for future designs to balance functionality, sustainability, and historical continuity, ensuring urban furniture contributes to cultural preservation and modern urban needs.

1. Introduction

Cities are dynamic and multifaceted systems that integrate communities with the built and natural environments, serving as platforms where biological needs, daily routines, and social interactions converge. As hubs of human activity across spatial and temporal dimensions, cities promote sustainable development by maintaining cultural continuity and adapting to contemporary needs [1]. According to Uslu [2], cities act as living witnesses of history, preserving traces of past civilizations and contributing to cultural continuity. Their spatial structure and identity are shaped by successive generations’ traditions, beliefs, lifestyles, and innovations, underscoring the interplay between human development and urban form.

The sustainability of cities extends beyond ecological concerns, encompassing the preservation and integration of cultural and historical values into urban development. Urban identity emerges from the interplay between the natural, built, and social environments, which collectively shape the distinctiveness and resilience of urban spaces. These identity-forming elements are rooted in historical continuity and are embedded within the collective urban memory [3,4]. Moreover, cities are in a perpetual transformation driven by political, environmental, and societal forces. Recognizing this evolution as an ongoing and dynamic process is essential for achieving sustainable urban development [5]. Within this context, urban furniture serves as a tangible medium through which memory, identity, and time converge, which is critical in enhancing the quality and continuity of urban experiences [6]. Integrating historically sensitive design approaches into urban elements can support cultural heritage preservation and socially inclusive urban sustainability. Urban furniture is defined as fixed equipment and structures used in open spaces within a city, typically serving various functions without specific users in mind. Urban furniture consists of emerging elements to meet city dwellers’ evolving and increasing needs [7]. Moreover, urban furniture is designed for functional purposes and is crucial in giving a city its unique identity. Similarly, Bekar et al. [8] emphasize that urban furniture is essential in enhancing urban aesthetics and shaping the identity of a city.

Within the broader context of sustainable urban development, urban furniture represents a critical component of inclusive, livable, and resilient cities. Particularly in historically significant areas, it must balance functionality, environmental responsibility, cultural continuity, and aesthetic integration. Effective design in these settings requires attention to sustainable materials, durability, ecological impact, and harmony with heritage values. Furthermore, urban furniture contributes to place-making, memory formation, and social cohesion—key pillars of sustainable urban life. The literature highlights that successful implementations prioritize ergonomic design; aesthetic coherence; long-term durability; and, increasingly, sustainability as a guiding principle in both design and placement.

The impact of urban furniture on the cultural identity of historical sites has been widely explored. Gjuroski [9] emphasizes that urban furniture contributes to the functional character of historical landscapes and facilitates inclusive design and social interaction, aligning with sustainability’s social and cultural pillars. Köksaldı and Turkan [10] analyzed urban furniture in Nicosia’s historical squares and found that most implementations lack harmony with the historical texture, both materially and aesthetically, undermining cultural continuity and usability. Similarly, Soffritti et al. [11] document the evolution of cast iron street furniture as both artistic and utilitarian elements, underscoring the importance of context-sensitive design. Integrating furniture into historical settings must consider both aesthetic harmony and sustainability. Scholars like Wan [12], Mumcu and Duzenli [13], and İlhan and Koç [14] argue for a design approach that respects the authenticity of heritage sites while ensuring usability and accessibility. However, many modern interventions still fail, compromising such spaces’ environmental and cultural integrity. Adding an emotional and psychological dimension to this discourse, Wang et al. [15] highlight the importance of emotionally resonant urban furniture. Using the Kano model, they demonstrate that incorporating traditional and culturally embedded elements significantly enhances user satisfaction and fosters emotional attachment to public spaces. Complementing this, Cheng et al. [16] employed ERP-based neuroscientific methods to show that perceptual inconsistencies—such as mismatches between design elements and user expectations—can trigger cognitive conflict, particularly in historical contexts. These findings emphasize the necessity of aligning urban furniture design with both the visual language of place and the cognitive-emotional expectations of users. Alongside form and emotion, the environmental sustainability of materials used in urban furniture is an increasingly prominent concern, particularly in heritage areas. Sipahi and Sipahi [17] highlight the ecological footprint of common materials such as concrete and polypropylene, advocating for biodegradable and renewable alternatives like natural wood. These material choices reduce carbon footprints and enhance visual compatibility with historical environments. Supporting this, Kou et al. [18] introduce a sustainability evaluation model that combines conservation principles with urban development, enabling a more holistic and long-term strategy for heritage management. Beyond materials, sustainable urban furniture must also address community needs, accessibility, and long-term usability. Akyol [19] and Wan [12] stress that durable, aesthetically compatible designs are vital for ensuring furniture longevity in historic districts. However, early approaches often neglected user-centered considerations, which are increasingly recognized as central to sustainable urban design. Innovative evaluation methods have been proposed to operationalize these goals. Gravagnuolo and Girard [20] developed a multicriteria decision-making framework that fosters inclusive and sustainable conservation, including heritage value, socio-economic factors, and stakeholder participation. Likewise, Tiboni et al. [21] utilize GIS-based tools to spatially assess urban regeneration efforts, offering data-driven insights into optimizing public space and infrastructure. These frameworks contribute to a broader understanding of sustainability, not just in environmental terms but through the lens of cultural continuity, social equity, and participatory governance.

Despite efforts to balance preservation and modernization, urban renewal projects often conflict with heritage conservation. Buyukkilic Kosun and Hamamcioglu Turan [22] examined urban transformation in historic districts near mosques in Turkey, revealing that modern interventions frequently disrupt the historical fabric and diminish urban identity. Similarly, Xia et al. [23] conducted a bibliometric analysis of global research trends in historic urban conservation, identifying digital documentation, risk assessment, and community engagement as critical research areas. These findings point to a crucial need for urban design strategies that align modern development with heritage protection, ensuring cultural sustainability and minimizing the erosion of authenticity.

A growing body of research has also focused on the ergonomics, functionality, and accessibility of urban furniture in historical settings. Studies by Gursoy [24] and Coban [25] highlight the importance of maintenance and functionality but often fail to address issues related to durability and inclusivity. Şişman and Gültürk [26] and Akın and Kavasoğulları [5] emphasize that urban furniture should be universally accessible, particularly for the elderly and disabled individuals. Their research proposes design solutions that enhance mobility and promote inclusivity in public spaces. In terms of historical texture and urban identity, studies by Güner [27] and Bekar et al. [8] analyze the role of urban furniture in shaping the aesthetic and identity of historic areas. However, these studies often overlook accessibility and perceptibility factors. More recent research by Akın and Kavasoğulları [5] and İlhan and Koç [14] provide more holistic frameworks by incorporating aesthetics, safety, material choice, ergonomics, accessibility, and sustainability as key design parameters. Despite this progress, further interdisciplinary research is needed to integrate emerging technologies with heritage preservation. Chen [28] calls for urban design approaches that include smart furniture and interactive systems, without compromising cultural integrity. In addition, several scholars stress the importance of participatory decision making to ensure that heritage spaces remain functional, inclusive, and socially relevant. The integration of AI, GIS technologies, and data-driven design offers novel ways to enhance sustainability in heritage sites while respecting historical authenticity.

Recent literature increasingly highlights the sustainability of materials, accessibility, and community-centered design as key parameters in heritage-sensitive urban furniture. However, emerging technologies, particularly artificial intelligence (AI), are reshaping how urban furniture is conceptualized, designed, and implemented. AI-generated urban furniture design introduces novel opportunities for addressing complex design parameters—such as historical context, ergonomic standards, sustainability, and public participation—through data-driven, iterative processes. Generative AI, in particular, has gained traction for its ability to produce novel, diverse design solutions through data-driven processes, acting as a co-creator in the ideation phase [29]. Studies show that AI tools enhance creative outcomes while reducing cognitive load in design students, suggesting strong potential for early-stage design support. In urban design, AI-generated images and tools like text-to-image synthesis have also explored visually compelling and historically sensitive design alternatives. These tools are being leveraged to visualize, prototype, and digitally fabricate urban elements, including street furniture, that harmonize with the architectural language of historic centers [30]. Yet, integrating AI in urban design, particularly in generating urban furniture, has sparked critical discussions around ethics, professional responsibility, and inclusivity in the planning process. AI-driven urban design, often framed within the urban artificial intelligence (UAI) concept, facilitates complex simulations and predictive modeling through digital twins, enhancing planners’ ability to evaluate design options at scale. However, while these tools offer novel efficiencies, scholars warn that UAI risks entrenching existing social inequities without ethical oversight if not designed with transparency and inclusivity [31]. As urban environments become increasingly saturated with AI-enabled technologies, such as smart benches or responsive public installations, the line between utility and surveillance grows thin. Chan [32] highlights that AI-powered smart cities may challenge traditional ethical frameworks by shifting decision-making power away from communities to opaque algorithmic processes, raising concerns about autonomy, data privacy, and moral agency. Bridging these technical advances with participatory design processes and co-production of knowledge becomes crucial in mitigating bias and ensuring community voices shape urban aesthetics and functionality.

Given these ethical and methodological considerations, exploring how AI-assisted design can be applied responsibly in real-world heritage contexts to support innovation and cultural preservation is essential. Designing urban furniture in historical environments through culturally sensitive and environmentally sustainable approaches plays a vital role in preserving urban identity and improving public spaces’ social and functional quality. This study aims to redesign the urban furniture in Sarahatun Mosque Square, located in the Harput historical site of Elazığ, to strengthen its connection to the historical context. By integrating AI-assisted design tools, this research proposes historically compatible urban furniture alternatives that align with the square’s aesthetic and architectural heritage. While previous studies have offered mainly qualitative assessments of existing urban furniture in heritage contexts, they often lack comprehensive, future-oriented design frameworks suitable for such sensitive settings. This study provides an original contribution by introducing a methodological framework that includes AI-assisted design generation and semantic differentiation analysis, validated through expert evaluations. This approach allows for an objective and multi-dimensional assessment of urban furniture, addressing historical compatibility, cultural symbolism, functionality, and user-centered sustainability. Ultimately, the research offers a novel perspective on urban identity and sustainable design in heritage areas. Bridging digital innovation with cultural preservation supports the creation of environmentally responsible and socially meaningful public spaces that contribute to the long-term sustainability of both cultural heritage and urban life.

2. Study Area

2.1. Harput Sarahatun Mosque Square

The study is situated in Harput, a historically rich settlement located approximately 5 km northeast of Elazığ city center, at coordinates 48°43′ N and 39°15′ E (Figure 1). With a continuous history of habitation dating back to 3000 BCE, Harput has served as a significant urban and trade center between the 16th and 19th centuries. The city’s historical and cultural landscape has been shaped by the layered influences of multiple pre-Islamic and post-Islamic civilizations, positioning Harput as a unique site for examining cultural sustainability. Harput’s strategic location and richly built environment led to its designation as an Urban Conservation Area on 14 December 1975. Further recognition came with Decision No. 1033, issued on 2 May 2007, by the Diyarbakır Regional Council for the Conservation of Cultural and Natural Assets, which classified the area as a historical urban conservation site, an urban site, and a first-degree archaeological site [33]. These designations underscore Harput’s role as a living cultural heritage site and an open-air museum, contributing significantly to sustainable tourism, heritage preservation, and community identity. Moreover, Harput was officially added to the UNESCO World Heritage Tentative List in 2018, reflecting its outstanding universal value. The process for full inscription on the World Heritage List is ongoing, further emphasizing the importance of protecting and enhancing its historical urban landscape. In this context, any intervention—including the design and placement of urban furniture—must reflect cultural sensitivity, aesthetic coherence, and long-term sustainability.

This study focuses on Sarahatun Mosque Square, a central gathering space within Harput that embodies functional and symbolic roles. Occupying 2556 m², the Sarahatun Mosque Square has been pedestrianized to reduce environmental impact and preserve the surrounding historical structures. It serves as a cultural tourism site and an essential urban space supporting social interaction, relaxation, learning, and even emergency preparedness—being designated as an emergency assembly area. The selection of this particular square as a case study is based on several critical factors:

• Efforts are underway for Harput to be inscribed on the UNESCO World Heritage Permanent List. In this context, the design of public spaces, including urban furniture, must reflect a high level of cultural sensitivity and compliance with international heritage standards.
• Sarahatun Mosque Square is located at the heart of Harput’s historic fabric. It functions as a key open public space with both religious and social significance, making it a critical site for studying context-aware urban furniture design in heritage environments.
• Similar challenges related to integrating contemporary urban furniture into historical settings are observed in many Turkish cities. The case of Harput serves as a representative model for addressing common issues of cultural continuity and aesthetic harmony in historic urban areas.
• The selected square offers an ideal testing ground for exploring how to balance preserving historical identity and contemporary urban life. It supports the development of design frameworks that are sustainably heritage-compatible.

2.2. Methodology

The methodology of the study consists of four stages:

• Determination of adjective pairs for visual perception with the Semantic Differentiation Method.
• Identify existing urban furniture photographs and AI-generated urban furniture designs for visual perception evaluation.
• Determination of survey measurement and data collection.
• Statistical analysis and assessment of the survey results.

2.2.1. Semantic Differentiation Method

Determining space-design perceptions is an experimental process. In this study, the Semantic Differentials Technique (SDT), developed by Osgood [34], was used to evaluate visual perception. This technique measures the associative meanings of concepts, personalities, or symbols in individuals’ minds. As widely used in environmental psychology, SDT analyzes user–environment interactions and examines subjective and environmental responses. Furthermore, it enables the semantic measurement of aesthetic perception based on evaluating a concept through pairs of opposing adjectives [35]. The visual evaluation assesses individuals’ or groups’ subjective perceptions of spaces. In other words, it aims to determine how different individuals perceive a concept. To evaluate visual perception, participants must rate a concept on a scale defined by two opposite adjectives (e.g., beautiful vs. ugly) [36]. This study developed a method to assess the compatibility of urban furniture designs with the historical fabric of Sarahatun Mosque Square, located in Harput-Elazığ. Central to this method was using adjective pairs, which formed the basis for evaluating how well urban furniture integrates with a heritage site’s spatial and cultural characteristics. The selection of adjective pairs followed a dual-criteria approach to ensure that they captured both spatial and historical dimensions of integration. First, a literature-based review was conducted using national and international studies that examined user perceptions of urban furniture in public and semi-public spaces such as squares, shopping malls, and campuses (

Table 1. Criteria used in the evaluation of urban furniture.

	Previous Studies	Akyol [19]	Wan [12]	Gursoy [24]	Coban [25]	Ghorab and Yücel Caymaz [37]	Guner [27]	Şişman and Gültürk [26]	Bekar et al. [8]	Gravagnuolo and Girard [20]	Mumcu and Duzenli [13]	Pavia [38]	Aydın Elmalı [6]	Gjuroski [9]	Soffritti et al. [11]	Akin and Kavasogullari [5]	Felek et al. [39]	Bingöl and Tezer [7]	Satiroglu et al. [40]	Kayahan [41]	Catalyurekoglu and Altiparmakogullari, [42]	Sarisin [43]	Varol [44]	Zamri et al. [45]	Chen [28]	Shehab El-Deen [46]	İlhan and Koc [14]	Sipahi and Sipahi [17]
Criteria
Material		+	+	+	+	-	+	+	-	-	+	-	+	-	+	+	-	-	+	-	-	-	+	-	-	-	-	+
Color		+	+	-	+	-	+	+	-	-	-	-	+	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Originality		+	-	+		+	-	+	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	+	-
Functionality		+	+	+	+	-	+	+	+	-	-	-	+	-	-	+	-	-	+	+	-	-	-	-	-	-	-	-
User Diversity/ Inclusive Design		+	+	-	-	-	-	-	-	-	+	+	-	+	-	+	-	-	-	-	+	-	-	-	-	-	+	-
Historical Texture/ Urban Identity/ Environmental Compatibility		-	+	-	-	+	-	-	-	-	+	+	-	+	+	+	-	-	-	+	+	+	-	+	+	+	+	-
Aesthetics		+	+	+	+	+	+	+	+	-	+	-	+	-	-	+	-	-	+	-	-	-	-	-	-	-	+	-
Ergonomics		+	-	+	+	-	-	-	-	-	-	-	-	+	-	+	+	-	-	-	+	-	-	-	-	-	+	-
Safety		+	+	-	-	-	-	-	-	-	-	-	-	-	-	+	-	-	-	-	-	-	-	-	-	-	+	-
Maintenance		+	+	+	+	-	-	+	-	-	-	-	-	-	-	+	-	-	+	-	-	-	+	-	-	+	+	-
Durability		+	-	+	-	-	-	-	-	-	-	-	-	-	-	-	-	-	+	-	-	-	+	-	+	-	+	-
Accessibility		-	+	-	-	-	-	-	-	-	-	-	-	-	-	+	-	-	-	-	-	-	-	-	-	-	-	-
Perceptibility		-	-	-	-	+	+	+	-	-	-	+	+	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Sustainability		-	+	-	-	-	-	-	-	-	-	-	+	-	-	-	-	-	+	-	-	-	-	-	+	-	-	-
Flexibility		+	+	-	-	-	-	-	-	-	-	-	-	+	-	-	-	-	-	-	+	-	-	-	-	-	-	-
Comfort		-	-	-	-	-	-	-	-	-	-	-	-	+	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Design		-	-	-	-	+	-	-	-	+	-	+	-	-	-	-	-	-	-	-	-	-	+	+	-	-	-	-

+ Evaluated criteria.

). These studies emphasized core evaluation dimensions, including aesthetic harmony, material appropriateness, cultural continuity, and contextual integration. Second, expert input was sought from professionals in architecture, landscape architecture, and cultural heritage to validate the contextual relevance of the selected adjective pairs. Expert feedback guided refinements to ensure that the pairs accurately reflected the unique character and environmental attributes of the Sarahatun Mosque Square.

Based on this two-tiered process, nineteen adjective pairs were identified and categorized into three overarching dimensions: Dominance, Excitement, and Pleasure. These primary dimensions were further divided into four sub-categories focused on specific qualities of urban furniture: (i) Functionality and Usability, (ii) Aesthetics and Visual Harmony, (iii) Compatibility with Historical Texture, and (iv) Material (

Table 2. Adjective pairs were used to evaluate Sarahatun Cami Meydanı (developed by Mahdavinejad and Abedi [36]).

Variable	Sub-Variable	Adjective Pair
Dominance: It is about a sense of personal freedom.	Functionality and Usability	Uncomfortable–Comfortable
Excitement: It refers to the presence and absence of exciting, interesting features related to the environment. It is mainly based on the design features of the spaces.	Aesthetics and Visual Harmony	Classical–Modern
	Historically Compatibility	Incompatible–Harmony
		Regional–Universal
		Rough–Elegant
		Boring–Interesting
		Non-Aesthetic–Aesthetic
		Ordinary–Original
		Unattractive–Attractive
		Complex–Simple
		Trivial–Glorious
	Material	Stone–Wood
		Metal–Wood

). This structure provided a comprehensive and context-sensitive framework for evaluating urban furniture’s visual, functional, and cultural resonance in a historically significant setting.

Using the semantic differentiation technique, a 7-point Likert scale was used to evaluate adjective pairs, with ratings ranging from −3 to +3. In this scale, negative values (−) were used to assess negative adjectives, while positive values (+) were used to evaluate positive adjectives. The degree of negativity for negative adjectives increased from −1 to −3, whereas the degree of positivity for positive adjectives increased from +1 to +3. The zero (0) value represents a neutral stance, indicating that the respondent perceives the concept as neither positive nor negative, reflecting an unbiased perspective.

2.2.2. Selection of Urban Furniture for Visual Perception

At this stage, images from two different sources were utilized for expert evaluation. The first source consisted of photographs taken in the field study, capturing the existing seating units, trash cans, and lighting elements in Sarahatun Mosque Square (Figure 2). To minimize perceptual variations caused by lighting differences, special care was taken to ensure that the photos were taken from similar angles on the same day and simultaneously. Additionally, images with minimal user presence were selected to enhance the clarity of the design perception.

As the second source, artificial intelligence (AI) tools—ChatGPT-4o, DALL·E 3, and Microsoft Copilot GCC (released in December 2024)—were used, reflecting the increasing role of AI in design applications in recent years. These tools were employed to generate urban furniture alternatives, marking a methodological integration of AI into historical urban design. Central to this process was the construction of input prompts, which guided the generation of seating units, trash cans, and lighting elements that embodied traditional and modern design elements. The aim was to ensure that the resulting furniture designs were aesthetically engaging and contextually compatible with historical environments.

The prompt keywords used in the AI-supported design process were crucial in ensuring that the generated outputs aligned with the cultural and spatial characteristics of Sarahatun Mosque Square. These keywords were derived through a comprehensive approach involving two primary sources: (1) the evaluation criteria commonly cited in the literature, as summarized in Table 1, and (2) expert opinions gathered during the initial phase of the research. Of particular importance in this context was the selection of materials. Prompt terms such as stone–wood and stone–concrete–wood combinations were intentionally included, as these material pairings were frequently emphasized in prior studies and recommended by experts for their compatibility with historical architectural contexts. Their incorporation into the AI prompts was intended to ensure environmental, cultural, and material continuity. From a design perspective, the prompts were deliberately constructed to include classical and modern stylistic features. This allowed for a comparative exploration of how different design approaches could interact with and respond to the historical urban fabric. The integration of this stylistic duality within the prompts supported the investigation of visual harmony and cultural coherence within the heritage setting. Furthermore, each AI-generated design was guided by the evaluative dimensions listed in Table 1, including functionality, aesthetics, historical compatibility, ergonomics, durability, and sustainability. These dimensions provided a comprehensive framework to ensure the generated designs were visually appropriate, practical, and culturally meaningful.

The preliminary evaluation of the images used in the survey was conducted with ten experts who have extensive knowledge of the Harput historical site and have previously undertaken scholarly research. Their deep familiarity with the local context was crucial in selecting AI-generated visuals. Specifically, the experts identified and selected those designs whose backgrounds closely reflected the architectural and spatial characteristics of the historical setting. From over 100 generated alternatives, 66 images were chosen for evaluation based on their contextual compatibility, as determined by expert consensus.

This careful selection process ensured that survey participants could assess the visual and architectural harmony of the urban furniture designs within an authentically grounded framework. Additionally, images devoid of human figures were intentionally selected to minimize distractions, allowing participants to focus on the urban furniture’s form, materials, and stylistic properties. As a result, the experts’ localized insights and context-sensitive judgments significantly strengthened the ecological and interpretive validity of the overall evaluation. Experts analyzed these selected AI-generated designs compared to the existing urban furniture and assessed their historical, cultural, and functional compatibility. Their comparative evaluation asked experts to develop urban furniture designs for seating units, trash cans, and lighting. The resulting design classifications and criteria established through this process are presented in Figure 3.

2.2.3. Data Collection

This study utilized web-based surveys (SurveyHero) to determine the compatibility levels of existing urban furniture in Sarahatun Mosque Square and the AI-generated urban furniture designs. Invitations were sent directly to experts through institutional email addresses, targeting professionals in architecture and landscape architecture with prior experience in urban design or cultural heritage. Expert selection criteria prioritized individuals with knowledge of historical environments and design expertise that could meaningfully contribute to the evaluation process. At this stage, the study aimed to reach Fırat University Faculty of Architecture faculty members, academics from other regional universities, and professionals working in the private sector and other public institutions. The primary objective was to identify the differences and similarities in spatial perception levels among individuals at different educational and professional stages. Therefore, experts specializing in historical sites or conducting research in this field were given priority in the selection process. One of the key objectives of this study was to analyze the similarities and differences in urban furniture evaluations among experts from different disciplines.

Accordingly, the survey consisted of two types of questions. The first section focused on the descriptive characteristics of an expert, covering five demographic and professional aspects: gender, area of expertise (e.g., architect, urban planner, and landscape architect), academic degree/title, institution, and professional experience. The second section included questions to evaluate the existing urban furniture and the AI-generated designs. At this stage, photographs of the existing urban furniture and AI-generated furniture designs for the square were assessed using the Semantic Differentiation Method based on predetermined adjective pairs. The survey form was created using an online platform, and the access link was shared with participants. A seven-point Likert scale (ranging from −3 to 3) was adapted into a numerical scale from 1 to 7 to facilitate a more precise evaluation of the adjective pairs. Participants reviewed photographs of each urban furniture design and rated their perception levels using this seven-point scale. In addition to this evaluation, participants were asked to select at least three of their most preferred AI-generated designs. This additional selection process helped determine which AI-generated design was perceived as the most compatible with the square, providing valuable insights for interpreting the study’s results.

Considering the potential impact of display-related variables on visual perception, a set of precautions was integrated into the survey administration process to ensure consistency across respondents. Given that the questionnaire was distributed and completed online, participants were explicitly instructed to conduct the evaluation using desktop or laptop devices with a minimum screen resolution of 1920 × 1080 pixels and to refrain from using mobile phones or tablets to minimize screen size and resolution discrepancies. Furthermore, they were advised to adjust their screen brightness to a neutral level and to complete the survey in a well-lit environment, thereby promoting standardized viewing conditions to the greatest extent possible. Notably, the expert participants represented a highly engaged and technically literate group. Their high level of compliance with the provided instructions is expected to have minimized perceptual variation, thus reducing the risk of bias introduced by device-related differences. As a result, the reliability and validity of the visual evaluation process were strengthened, supporting the credibility of the study’s findings.

2.2.4. Data Analysis

Expert responses were analyzed using IBM SPSS Statistics 22, provided by Fırat University. The statistical methods employed in the evaluation of the results consisted of two stages:

Descriptive statistics: Firstly, both existing urban furniture and AI-assisted designs were evaluated by an expert group using the semantic differentiation method to determine their compatibility with the historical texture. Each pair of adjectives identified in the process was rated to assess perceived integration with historical surroundings. Additionally, descriptive statistical methods were employed to analyze the relationship between Sarahatun Mosque Square’s perceived attractiveness and the participants’ socio-demographic characteristics. In this phase,

• Frequency analysis was conducted to examine the general distribution of the data.
• Means and standard deviations were calculated to determine the variables’ central tendency and dispersion measures.
• Graphical representations (e.g., bar charts and histograms) were used to visualize the findings effectively.

This approach provided a comprehensive statistical foundation for evaluating the perception of urban furniture in historical settings and understanding how different design approaches align with expert opinions.

Explanatory Factor Analysis (EFA): Factor analysis was conducted to determine the effectiveness of adjective pairs in evaluating the furniture designs generated by the AI-based model. This statistical method reduces a large set of variables into a smaller number of meaningful factors that explain variations within the dataset [47]. It is an analytical technique widely applied to structure perceptual data, mainly aesthetic evaluations. The primary objective of this study was to identify and group adjective pairs based on their influence on visual perception. Factor analysis helps detect data patterns by clustering correlated variables under common factors. This study categorized adjective pairs into distinct factor groups, each representing a specific dimension of furniture perception. An essential consideration in factor analysis is the adequacy of the sample size and the reliability of responses. A larger sample size increases the robustness of the analysis, while a high number of variables may affect the accuracy of factor extraction [47]. Given the presence of 12 adjective pairs in this study, appropriate adjustments were made to maintain analytical accuracy and ensure reliable results.

3. Results

This study conducted web-based surveys between December 2024 and February 2025. A total of 35 experts participated; however, responses from four participants were excluded due to incomplete answers or failure to complete the survey. As a result, the visual assessment of Sarahatun Mosque Square was based on the responses of 31 experts.

3.1. Data on Experts

Firstly, the descriptive statistics of the expert group evaluating the historical compatibility of existing and AI-generated urban furniture designs were presented (

Table 3. Descriptive characteristics of experts.

		Landscape Architecture (n = 16)		Architecture (n = 15)		Total (n = 31)
Socio-Demographic Characteristics		The Number of Participants	Percentage	The Number of Participants	Percentage	The Number of Participants	Percentage
		n	%	n	%	n	%
Gender
	Men	9	56.2	3	20.0	12	38.7
	Women	7	43.8	12	80.0	19	61.3
Professional Experience
	0–5 years	2	12.5	8	53.4	10	32.3
	6–10 years	2	12.5	-	0	2	6.5
	11–15 years	6	37.5	3	20.0	9	29.0
	16–20 years	2	12.5	2	13.3	4	12.9
	Over 20 years	4	25.0	2	13.3	6	19.4
Academic Degree/Title
	Graduated	1	6.25	6	40.0	7	22.6
	Master	3	18.75	2	13.3	5	16.1
	Ph.D.	1	6.25	2	13.3	3	9.7
	Assistant professor	3	18.75	1	6.7	4	12.9
	Associate professor	4	25.0	3	20.0	7	22.6
	Professor	4	25.0	1	6.7	5	16.1
Institution
	University	14	87.5	7	46.7	21	67.7
	Public sector	1	6.25	2	13.3	3	9.7
	Private sector	-	-	4	26.7	4	12.9
	Other	1	6.25	2	13.3	3	9.7

). The study was conducted with experts at the national level, incorporating participants from various national universities, private sector organizations, and public institutions. This diverse institutional representation enhances the credibility and applicability of the findings across different professional and academic perspectives. A total of 21 experts from 9 universities participated in the study, representing Çukurova University, Fırat University, İnönü University, Ege University, Yüzüncü Yıl University, İskenderun Technical University, Kahramanmaraş Sütçü İmam University, Niğde Ömer Halisdemir University, and Selçuk University. Additionally, 10 experts from public institutions and private sector organizations contributed, bringing practical, industry-driven insights to complement the academic perspectives.

The expert’s descriptive characteristics analysis reveals a diverse yet academically inclined group. The majority of the participants were female (61.3%), with women being particularly dominant in the fields of Architecture (80.0%), while Landscape Architecture had a relatively higher proportion of male participants (56.3%). In terms of professional experience, a significant portion of the experts were either early-career professionals (32.3% with 0–5 years of experience) or mid-career professionals (29.0% with 11–15 years of experience), with fewer participants in the 6–10 years and 16–20 years categories. However, a notable 19.4% had over 20 years of experience, indicating the presence of highly seasoned professionals in the study. Regarding academic qualifications, most participants held advanced degrees, with 22.6% being associate professors and 16.1% being professors, demonstrating a strong academic foundation among the experts. Bachelor’s degree holders (22.6%) were mainly from the Architecture field, suggesting that this discipline had a relatively higher proportion of professionals still in the early stages of their academic or professional careers. The institutional affiliation data further supports the strong academic presence, as 67.7% of participants were university-affiliated, while only 12.9% were from the private sector, mainly from Architecture, and a smaller portion (9.7%) worked in the public sector. Overall, the findings indicate that the participant pool was predominantly composed of experienced academics and professionals from Landscape Architecture and Architecture, with a strong representation from universities. The distribution of gender, professional experience, and academic titles suggests a balanced mix of senior and junior experts, contributing a wide range of perspectives to the study (Table 3).

3.2. Perception of Urban Furniture in the Context of Semantic Differentiation Method

Secondly, experts’ perceptions regarding existing and AI-assisted urban furniture designs were evaluated using a seven-point Likert scale based on adjective pairs determined by the Semantic Differential Technique. The study identified key perceptual contrasts by analyzing the semantic differentiation of existing and AI-generated urban furniture, highlighting which design elements contribute to a stronger sense of historical integration, aesthetic appeal, and functional effectiveness (Figure 4). The findings derived from expert opinions are as follows:

The comparative analysis of urban furniture perception highlights significant differences between modern, traditional, and existing furniture based on comfort, aesthetics, usability, and cultural relevance. The existing urban furniture was rated as moderately uncomfortable (mean = 3.39, SD = 1.52), somewhat incompatible (mean = 3.45, SD = 1.82), and less aesthetic (mean = 2.55, SD = 1.59), suggesting a need for design improvements. In contrast, modern furniture elements were perceived as significantly more comfortable, with benches (mean = 5.68, SD = 1.30), trash cans (mean = 5.74, SD = 1.18), and lighting (mean = 5.84, SD = 1.10) receiving the highest comfort ratings. Additionally, modern designs were viewed as highly modern (lighting mean = 6.13, SD = 1.02), harmonious (trash can mean = 5.68, SD = 1.35), and aesthetically pleasing (bench mean = 6.13, SD = 0.88). These findings indicate that modern furniture is preferred for its functionality and visual appeal, making it an ideal choice for historical urban spaces.

On the other hand, traditional furniture was rated lower in modernity but higher in historical and cultural significance. Traditional elements were seen as less comfortable than modern designs, with benches (mean = 4.23, SD = 1.75), trash cans (mean = 3.87, SD = 1.67), and lighting (mean = 4.74, SD = 1.46) receiving moderate comfort ratings. However, traditional furniture was appreciated for its simplicity and elegance, with trash cans (mean = 6.39, SD = 0.95) and benches (mean = 6.16, SD = 1.24) being perceived as the simplest. Additionally, traditional furniture was considered highly culturally significant, as seen in the low “trivial” scores for benches (mean = 2.52, SD = 1.67) and lighting (mean = 2.61, SD = 1.17), reinforcing their role in maintaining the authenticity of historical urban spaces.

According to the results, modern furniture excels in comfort, aesthetics, and adaptability, making it well-suited for contemporary urban environments. In contrast, traditional furniture retains value through cultural significance, simplicity, and historical harmony. A balanced integration of modern and traditional elements in urban furniture design could help satisfy users’ needs while preserving the character of historical spaces.

Experts evaluated urban furniture designs under two main categories, wood–stone/concrete and wood–metal, to assess material preferences in Sarahatun Mosque Square. The analysis revealed a neutral stance between these material combinations, suggesting that both are considered equally functional and acceptable within the urban environment, with users prioritizing aesthetics and functionality over material choices. The evaluation of existing urban furniture showed that benches made of wood–concrete (mean = 3.32, SD = 2.04) and trash cans made of wood–metal (mean = 4.06, SD = 2.13) received moderate ratings. When comparing AI-assisted urban furniture, modern benches (mean = 3.35, SD = 2.24) were rated lower than classical benches (mean = 4.29, SD = 2.31), while modern lighting (mean = 4.35, SD = 2.09) outperformed classical lighting (mean = 3.94, SD = 2.24). Similarly, modern trash cans (mean = 3.45, SD = 2.06) were rated lower than classical trash cans (mean = 4.74, SD = 1.88), indicating a slight preference for classical furniture elements, particularly benches and trash cans. In contrast, modern lighting was perceived as more suitable for public spaces. These findings suggest that a balanced approach, incorporating elements of both contemporary and classical designs, may be optimal for achieving aesthetic appeal and functionality in urban furniture planning.

3.3. Explanatory Factor Analysis

Thirdly, Explanatory Factor Analysis (EFA) was used to determine which adjective pairs influence expert perception regarding the existing and AI-supported urban furniture designs in Harput Sarahatun Mosque Square. The factors affecting expert perception were identified through this analysis. Conducting EFA requires attention to three key points. First, only factors with an initial eigenvalue greater than 1.00 were considered. Second, adjective pairs with a factor loading below 0.50 were excluded from the factor groups. Finally, adjective pairs that overlapped under multiple factors were removed from the evaluation, and the factor analysis was repeated to refine the identified factors [48,49].

The adequacy of the sample size for the analysis was tested using the Kaiser–Meyer–Olkin (KMO) measure. According to this criterion, a KMO value above 0.60 indicates that the sample size is sufficient for performing EFA. The analysis showed that the obtained KMO values exceeded 0.60, confirming that the sample size was appropriate for factor analysis. The factor-naming process was based on the methodology proposed by Yakin İnan and Özdemir Sönmez [50]. The study identified two factors in the evaluation of existing urban furniture (

Table 4. EFA results for existing urban furniture in Harput Sarahatun Mosque Square.

Existing Urban Furniture in Sarahatun Mosque Square
Factor	Factor Loading	Factor	Factor Loading
Factor 1. Aesthetic and Visual Appeal a Eigenvalue = 4.70 Explained variance = 36.17%		Factor 2. Symbolic and Structural Simplicity a Eigenvalue = 1.99 Explained variance = 15.37%
Ordinary–Original	0.96	Complex–Simple	0.82
Boring–Interesting	0.91	Trivial–Glorious	0.72
Non-Aesthetic–Aesthetic	0.90	Material for trash can	0.63
Rough–Elegant	0.78
Unattractive–Attractive	0.64
Factor 3. Historical Compatibility a Eigenvalue = 1.62 Explained variance = 12.48%		Factor 4. Material a Eigenvalue = 1.27 Explained variance = 9.77%
Regional–Universal	0.87	Material selection for benches	0.97
Incompatible–Harmony	0.59
Classical–Modern	0.52
		Total Explained Variance (%)	73.80
a KMO = 0.63 and Barlett’s Test of Sphericity = 0.000
Cross-loaded factors: None Factor loading less than 0.50: Uncomfortable–Comfortable (0.42)

), two factors in the evaluation of modern urban furniture (

Table 5. EFA results for AI-generated modern-style urban furniture.

AI-Generated Modern Style Urban Furniture
Benches		Trash Can		Lighting
Factor	Factor Loading	Factor	Factor Loading	Factor	Factor Loading
Factor 1. Historical Compatibility a Eigenvalue = 5.56 Explained variance = 46.35%		Factor 1. Historical Compatibility a Eigenvalue = 5.58 Explained variance = 55.83%		Factor 1. Historical Compatibility a Eigenvalue = 7.47 Explained variance = 62.30%
Non-Aesthetic–Aesthetic	0.94	Boring–Interesting	0.95	Boring–Interesting	0.95
Boring–Interesting	0.90	Non-Aesthetic–Aesthetic	0.91	Rough–Elegant	0.91
Unattractive–Attractive	0.87	Unattractive–Attractive	0.88	Uncomfortable–Comfortable	0.89
Rough–Elegant	0.84	Rough–Elegant	0.86	Ordinary–Original	0.87
Classical–Modern	0.83	Ordinary–Original	0.86	Non-Aesthetic–Aesthetic	0.85
Regional–Universal	0.65	Classical–Modern	0.83	Classical–Modern	0.84
Uncomfortable–Comfortable	0.62	Incompatible-Harmony	0.68	Regional–Universal	0.81
Ordinary–Original	0.59	Uncomfortable–Comfortable	0.66	Incompatible–Harmony	0.78
Incompatible–Harmony	0.53			Unattractive–Attractive	0.74
				Trivial–Glorious	0.70
Factor 2. Material and simplicity a Eigenvalue = 1.67 Explained variance = 13.96%		Factor 2. Material and simplicity a Eigenvalue = 1.28 Explained variance = 12.83%		Factor 2. Material and simplicity a Eigenvalue = 1.20 Explained variance = 10.01%
Material selection	0.81	Complex–Simple	0.89	Material selection	0.87
Complex–Simple	0.74	Material selection	0.52	Complex–Simple	0.57
Total Explained Variance (%)	60.31	Total Explained Variance (%)	68.67	Total Explained Variance (%)	72.31
a KMO = 0.84 Barlett’s Test of Sphericity = 0.000		a KMO = 0.74 Barlett’s Test of Sphericity = 0.000		a KMO = 0.88 Barlett’s Test of Sphericity = 0.000
Cross-loaded factors: None Factor loading less than 0.50: Trivial–Glorious (0.38)		Cross-loaded factors: Regional–Universal, Trivial–Glorious Factor loading less than 0.50: None		Cross-loaded factors: None Factor loading less than 0.50: None

), and two to three factors in the assessment of classical urban furniture (

Table 6. EFA results for AI-generated classical-style urban furniture.

AI-Generated Classical Style Urban Furniture
Benches		Trash Can		Lighting
Factor	Factor Loading	Factor	Factor Loading	Factor	Factor Loading
Factor 1. Visual appeal a Eigenvalue = 4.58 Explained variance = 45.77%		Factor 1. Historical Compatibility a Eigenvalue = 4.89 Explained variance = 44.48%		Factor 1. Historical Compatibility a Eigenvalue = 5.67 Explained variance = 56.73%
Non-Aesthetic–Aesthetic	0.94	Unattractive–Attractive	0.90	Boring–Interesting	0.93
Boring–Interesting	0.94	Non-Aesthetic–Aesthetic	0.89	Non-Aesthetic–Aesthetic	0.91
Unattractive–Attractive	0.90	Boring–Interesting	0.88	Unattractive–Attractive	0.90
Rough–Elegant	0.86	Ordinary–Original	0.78	Incompatible–Harmony	0.88
Incompatible–Harmony	0.78	Rough–Elegant	0.78	Rough–Elegant	0.84
Uncomfortable–Comfortable	0.71	Incompatible–Harmony	0.74	Uncomfortable–Comfortable	0.79
		Uncomfortable–Comfortable	0.73	Ordinary–Original	0.73
Factor 2. Design perception a Eigenvalue = 2.15 Explained variance = 21.50%		Factor 2. Design perception a Eigenvalue = 1.59 Explained variance = 14.45%		Factor 2. Material and simplicity a Eigenvalue = 1.52 Explained variance = 15.22%
Trivial–Glorious (0.38)	0.83	Trivial–Glorious	0.72	Classical–Modern	0.88
Complex–Simple	0.80	Complex–Simple	0.67	Trivial–Glorious	0.83
Classical–Modern	0.73
Factor 3. Cultural scope a Eigenvalue = 1.07 Explained variance = 10.74%		Factor 3. Cultural scope a Eigenvalue = 1.35 Explained variance = 12.29%
Regional–Universal	0.94	Classical–Modern	0.78
		Regional–Universal	0.73
Total Explained Variance (%)	78.01	Total Explained Variance (%)	71.23	Total Explained Variance (%)	71.95
a KMO = 0.73 Barlett’s Test of Sphericity = 0.000		a KMO = 0.77 Barlett’s Test of Sphericity = 0.000		a KMO = 0.79 Barlett’s Test of Sphericity = 0.000
Cross-loaded factors: Ordinary–Original, Material selection Factor loading less than 0.50: None		Cross-loaded factors: Material selection Factor loading less than 0.50: None		Cross-loaded factors: None Factor loading less than 0.50: Material selection (0.33)

). The naming of factors was guided by the characteristics of the adjective pairs, ensuring that the factor groups reflected the conceptual diversity of the associated adjectives.

The results of the EFA indicate that different adjective pairs influenced expert perception in varying degrees. Not all 12 adjective pairs effectively determined the factors (Table 4, Table 5 and Table 6). Additionally, the EFA results suggest that urban furniture exhibits different levels of compatibility with the historical texture, further highlighting variations in expert evaluations. Accordingly, the findings obtained from the analysis are as follows:

Existing furniture design: EFA results identified four distinct factors that influence the perception of urban furniture in the existing situation. The total explained variance of the analysis was 73.80%, indicating a strong explanatory power. The Kaiser–Meyer–Olkin (KMO) value of 0.63 and Bartlett’s Test of Sphericity (p = 0.000) confirmed that the sample was suitable for factor analysis. No cross-loaded factors were found, and only one adjective pair (Uncomfortable–Comfortable) had a factor loading below the 0.50 threshold, leading to its exclusion from the factor structure (Table 4).

The first factor, labeled “Aesthetic and Visual Appeal”, had an eigenvalue of 4.70 and accounted for 36.17% of the total variance. This factor primarily comprised aesthetic perception, attractiveness, and visual harmony variables. The highest factor loading was observed in Ordinary–Original (0.96), followed by Boring–Interesting (0.91) and Non-Aesthetic–Aesthetic (0.90). Other significant contributors to this factor were Rough–Elegant (0.78) and Unattractive–Attractive (0.64), indicating that the perceived aesthetic quality of urban furniture plays a crucial role in shaping expert evaluations.

The second factor, “Symbolic and Structural Simplicity”, had an eigenvalue of 1.99, explaining 15.37% of the total variance. This factor was associated with urban furniture elements’ simplicity and symbolic meaning. Complex–Simple (0.82) had the highest loading, suggesting that design clarity and ease of interpretation strongly influence perceptions. The factor also included Trivial–Glorious (0.72) and Material for Trash Can (0.63), indicating that symbolic representation and material selection contribute to the overall perception of simplicity in urban furniture design.

The third factor, “Historical Compatibility”, had an eigenvalue of 1.62, accounting for 12.48% of the total variance. This factor captured the degree to which urban furniture aligns with historical and cultural contexts. The highest loading was found in Regional–Universal (0.87), emphasizing whether the furniture is perceived as specific to a region or adaptable to a broader cultural context. Other contributing variables included Incompatible–Harmony (0.59) and Classical–Modern (0.52), reflecting the extent to which urban furniture blends with or disrupts the historical and architectural integrity of the surroundings. The fourth factor, “Material Selection”, had an eigenvalue of 1.27 and explained 9.77% of the total variance. This factor specifically addressed material preferences in urban furniture design. Material selection for benches (0.97) was the dominant variable in this factor, underscoring the importance of material choices in influencing perceptions of durability, sustainability, and usability.

Modern urban furniture: The EFA result showed that the total explained variance was 60.31% for benches, 68.67% for trash cans, and 72.31% for lighting elements, indicating a strong explanatory power of the identified factors. The Kaiser–Meyer–Olkin (KMO) values were 0.84 for benches, 0.74 for trash cans, and 0.88 for lighting elements, confirming that the sample size was sufficient for factor analysis. Bartlett’s Test of Sphericity was significant (p = 0.000) in all cases, supporting the appropriateness of the factor analysis model. Additionally, no cross-loading issues were found in benches and lighting elements. However, for trash cans, the “Regional–Universal” and “Trivial–Glorious” pairs exhibited cross-loading, indicating some overlap in perceptions regarding historical integration and grandeur (Table 5).

EFA for modern-style urban furniture, including benches, trash cans, and lighting elements, indicates a two-factor structure: “Historical Compatibility” and “Material and Simplicity”. The first factor, Historical Compatibility, emerged as the dominant factor across all three furniture types, explaining 46.35% of the variance for benches, 55.83% for trash cans, and 62.30% for lighting elements. The eigenvalues ranged from 5.56 to 7.47, highlighting the strong influence of this factor in expert evaluations. The adjective pairs “Non-Aesthetic–Aesthetic”, “Boring–Interesting”, “Unattractive–Attractive”, and “Rough–Elegant” had the highest loadings across all categories, indicating that perceptions of aesthetic quality and attractiveness significantly impact historical compatibility in modern furniture design. “Classical–Modern” and “Regional–Universal” were also influential, particularly in lighting elements (loading = 0.84 and 0.81, respectively), suggesting that the integration of modern aesthetics with historical settings remains a critical factor. The “Uncomfortable–Comfortable” adjective pair was included in this factor for benches (loading = 0.62) and trash cans (loading = 0.66), demonstrating that comfort is perceived as part of historical harmony in urban furniture evaluations. The “Incompatible–Harmony” variable was relevant in trash cans (0.68) and lighting elements (0.78), reinforcing the importance of blending with the historical environment. The “Trivial–Glorious” pair had a low factor loading in benches (0.38) and was removed from the final factor structure.

The second factor, Material and Simplicity, accounted for 13.96% of the variance for benches, 12.83% for trash cans, and 10.01% for lighting elements, with eigenvalues ranging from 1.20 to 1.67. This factor primarily reflects preferences related to material selection and simplicity of design. “Material Selection” had a strong influence on benches (0.81), trash cans (0.52), and lighting elements (0.87), indicating that the choice of materials plays a crucial role in urban furniture design preferences. “Complex–Simple” was also a key variable, loading significantly in all three categories (benches = 0.74, trash cans = 0.89, lighting = 0.57), suggesting that simplicity in design is generally preferred over complexity in modern-style urban furniture.

Classical urban furniture: The results of the explanatory factor analysis (EFA) conducted on AI-generated classical-style urban furniture reveal a three-factor structure for benches and trash cans, while a two-factor structure emerged for lighting elements. The total explained variance for benches was 78.01%, for trash cans 71.23%, and for lighting elements 71.95%, indicating that the identified factors effectively account for the variation in expert evaluations. The Kaiser–Meyer–Olkin (KMO) values for all three categories were above 0.70, and Bartlett’s test of sphericity was significant (p < 0.001), confirming the suitability of the data for factor analysis (Table 6).

Factor 1: Visual Appeal had the highest eigenvalue (4.58) for benches and explained 45.77% of the variance. This factor included attributes related to aesthetic perception, such as Non-Aesthetic–Aesthetic (0.94), Boring–Interesting (0.94), Unattractive–Attractive (0.90), Rough–Elegant (0.86), Incompatible–Harmony (0.78), and Uncomfortable–Comfortable (0.71). Factor 2: Design Perception accounted for 21.50% of the variance and included descriptors related to perceived design complexity and grandeur, such as Trivial–Glorious (0.83), Complex–Simple (0.80), and Classical–Modern (0.73). Factor 3: Cultural Scope, explaining 10.74% of the variance, included the attribute Regional–Universal (0.94).

For trash cans, Factor 1: Historical Compatibility had an eigenvalue of 4.89, explaining 44.48% of the variance, and included variables related to perceived attractiveness and historical integration, such as Unattractive–Attractive (0.90), Non-Aesthetic–Aesthetic (0.89), Boring–Interesting (0.88), Ordinary–Original (0.78), Rough–Elegant (0.78), Incompatible–Harmony (0.74), and Uncomfortable–Comfortable (0.73). Factor 2: Design Perception, accounting for 14.45% of the variance, included Trivial–Glorious (0.72) and Complex–Simple (0.67). Factor 3: Cultural Scope, explaining 12.29% of the variance, included Classical–Modern (0.78) and Regional–Universal (0.73).

For lighting elements, Factor 1: Historical Compatibility had the highest eigenvalue (5.67) and explained 56.73% of the variance, comprising aesthetic and integration-related attributes, such as Boring–Interesting (0.93), Non-Aesthetic–Aesthetic (0.91), Unattractive–Attractive (0.90), Incompatible–Harmony (0.88), Rough–Elegant (0.84), Uncomfortable–Comfortable (0.79), and Ordinary–Original (0.73). Factor 2: Material and Simplicity, explaining 15.22% of the variance, including attributes associated with material selection and overall design style, such as Classical–Modern (0.88) and Trivial–Glorious (0.83).

Some variables, such as material selection and Ordinary–Original, displayed cross-loading effects in the benches and trash cans categories, suggesting that these descriptors may influence multiple perceptual dimensions. In contrast, lighting elements did not exhibit cross-loading effects, and all included variables had factor loadings above the 0.50 threshold.

The results of the EFA indicate that the perception of urban furniture varies significantly depending on its design style (existing, modern, or classical) and its alignment with aesthetic, historical, and functional attributes. The identified factors demonstrate that classical-style urban furniture is predominantly evaluated based on historical compatibility and cultural integration. In contrast, modern and existing urban furniture is assessed through aesthetic appeal, material selection, and simplicity. AI-generated classical furniture exhibits a more structured factor differentiation, with a higher total explained variance, suggesting that these designs evoke stronger and more consistent expert evaluations. Additionally, cross-loading effects observed in some existing and modern furniture categories indicate that specific design attributes, such as material selection and originality, influence multiple perceptual dimensions simultaneously. Overall, the EFA findings highlight the importance of harmonizing visual appeal, historical authenticity, and material sustainability in urban furniture design, emphasizing that aesthetic and contextual factors shape the perception of public space infrastructure.

3.4. Historical Compatibility of Urban Furniture

Finally, experts were asked to select the three AI-generated urban furniture designs—benches, trash cans, and lighting elements—that they deemed most compatible with the historical texture. The findings reveal that existing urban furniture received the lowest ratings regarding historical compatibility, indicating a significant disconnect between the current urban furniture and the historical identity of Sarahatun Mosque Square. Following this, classical-style furniture was rated moderately compatible. In contrast, modern urban furniture designs emerged as the most preferred by experts, suggesting that they were perceived as the most aesthetically and functionally suitable for integration into the historical setting. Although the Explanatory Factor Analysis (EFA) identified the key factors influencing historical compatibility, it did not indicate the most suitable urban furniture designs. Given the misalignment of existing furniture with the historical fabric of the site, experts evaluated alternative urban furniture designs to determine the most appropriate options for enhancing the historical and aesthetic integrity of Sarahatun Mosque Square (Figure 5). These findings emphasize the importance of context-sensitive design strategies in preserving historical urban landscapes while incorporating contemporary design solutions. Moreover, they also offer insights into preferred styles, materials, and forms that contribute to a cohesive urban identity within historical settings.

• Modern Bench Designs: Among modern bench designs, the most highly rated were ID-9 (41.9%), ID-1, and ID-4 (38.7%), which were preferred due to their strong alignment with key aesthetic attributes such as aesthetic appeal (mean = 6.13), modernity (mean = 6.23), and attractiveness (mean = 6.00). Conversely, ID-3 and ID-12 (16.1%) were rated as the least preferred, likely due to lower evaluations in showiness (mean = 4.65) and originality (mean = 5.32), suggesting that these designs lacked distinctiveness and impact.
• Classical Bench Designs: ID-11 (51.6%) and ID-5 (41.9%) received the highest preference ratings in classical bench designs. Unlike modern designs, classical benches were evaluated with more variability, mainly influenced by showiness (mean = 6.16) and traditionalism (mean = 2.81). The least favored classical bench designs were ID-8 (6.5%) and ID-6 (9.7%), which were negatively associated with attributes such as roughness (mean = 3.74) and complexity (mean = 2.52), indicating a perception of excessive ornamentation or impracticality.
• Modern Trash Can Designs: For modern trash cans, designs ID-7 (54.8%) and ID-5 (51.6%) were identified as the most suitable, aligning with modernity (mean = 6.00) and comfort (mean = 5.74). These designs were considered ergonomic and environmentally friendly, contributing to their higher preference ratings. On the other hand, ID-4 (9.7%) and ID-3 (22.6%) were the least favored. Despite being associated with comfort (mean = 5.74), experts identified ergonomic concerns that reduced their appeal. Additionally, they were rated more straightforward (mean = 3.74) and less visually engaging (mean = 5.65), contributing to their lower preference scores.
• Classical Trash Can Designs: Among classical trash cans, ID-1 (58.1%) and ID-3 (54.8%) were the most preferred. Their selection was influenced by their alignment with traditional (mean = 2.87), local (mean = 3.61), and showy (mean = 6.39) attributes, making them visually fitting for the historical setting. Conversely, ID-6 (25.8%) and ID-2 (29.0%) were rated the least preferred due to their perceived discomfort (mean = 3.87), roughness (mean = 3.65), and complexity (mean = 2.65), making them appear less refined and more intrusive in the environment.
• Modern Lighting Designs: Experts strongly preferred modern lighting ID-2 (58.1%), followed by designs ID-1, ID-4, and ID-5 (45.2%), which were equally favored. These designs were highly rated due to their modern (mean = 6.13), universal (mean = 6.06), and aesthetic (mean = 5.77) qualities, making them the most suitable for public spaces. In contrast, ID-7 (19.4%) and ID-8 (22.6%) were rated the least preferred, as they were perceived as too simple (mean = 5.29), reducing their impact and distinctiveness.
• Classical Lighting Designs: In classical lighting designs, ID-8 (71.0%) was the most highly rated, followed by ID-7 (51.6%), indicating strong alignment with historical aesthetics. These designs were positively associated with traditional (mean = 2.90), showy (mean = 5.81), and inviting (mean = 5.06) characteristics, making them well-suited for the historical setting. On the other hand, ID-2 (9.7%) and ID-3 (22.6%) were the least preferred. Their lower preference was attributed to being complex (mean = 2.61) and lacking elegance (mean = 4.42), making them appear rougher and less refined.

The expert evaluations indicate that modern and classical urban furniture designs can suit historical environments, depending on their aesthetic compatibility, functionality, and ergonomic considerations. Modern lighting elements were preferred over classical ones, while classical benches and trash cans were slightly more favored than modern ones. The findings suggest that a balanced approach integrating modern functionality and historical authenticity would be the most effective strategy for designing urban furniture in culturally significant locations like Harput Sarahatun Mosque Square.

4. Discussion

4.1. Methodological Aspects

Previous studies on urban furniture in historical contexts emphasize different aspects such as aesthetics, material selection, sustainability, and historical integration. The results reveal the importance of developing sustainable and user-friendly designs sensitive to historical and cultural contexts [46]. Research on design identity and cultural impact [9,11] primarily focuses on visual identity and material analysis without structured historical assessments. Fallacara et al. [30] further demonstrate the viability of AI-generated visual outputs for heritage-sensitive urban furniture through digital fabrication, reinforcing the feasibility of our design generation approach. Our study systematically integrates historical, aesthetic, and material dimensions, bridging the gap between AI-generated and existing urban furniture evaluations. Studies using multicriteria decision-making frameworks [20] and analyses of urban transformation [22] highlight the impact of modern interventions but lack direct design evaluations. Our research expands on these by incorporating semantic differentiation and EFA to identify key factors affecting historical compatibility. Sustainability-focused studies [17,18] assess environmental impacts but do not address urban furniture’s perceptual and historical alignment, which our study directly evaluates through expert analysis. Similarly, research on phenomenology and emotional design [38] considers user experience but does not integrate historical material suitability. Wang et al. [15] further reinforce this by showing that emotional attachment increases when traditional cultural elements are used in street furniture, especially in heritage contexts. Our study contributes by combining AI-generated evaluations with expert-driven assessments for a data-driven yet historically informed approach. Recent neurocognitive research by Cheng et al. [16] has also highlighted that inconsistent aesthetic stimuli in historical urban settings cause cognitive dissonance and reduce perceptual congruence. Our findings align with this by showing that designs perceived as historically compatible had higher expert approval, reinforcing the importance of perceptual harmony.

Our study introduces a novel AI-supported approach that integrates historical compatibility and expert perception, distinguishing itself through EFA’s structured and comparative evaluation. The novelty of our research lies in its integration of AI-generated urban furniture designs with expert assessment, allowing for a comparative framework that assesses both AI-generated and traditional designs. This approach bridges historical authenticity, modern computational design, and expert perception, offering urban planners and designers a more adaptable evaluation model. Our findings reveal that modern designs favor aesthetics and material innovation, while classical designs align more with historical contexts. The EFA results indicate that expert perception is shaped by visual appeal, historical compatibility, and material selection, distinguishing our study from others that assess urban furniture based on standalone design principles [44,45]. Our study presents a comprehensive, AI-assisted, and expert-validated evaluation framework for urban furniture in historical settings. By integrating factor analysis, semantic differentiation, and expert perception metrics, we provide a multi-dimensional approach that assesses historical compatibility and informs future urban design strategies with data-driven insights.

4.2. Similarities in Historical Compatibility Evaluations

One of the most consistent findings across studies is the importance of visual and material harmony in historical urban spaces. Previous studies emphasize that designs should align with the local context to maintain authenticity and strengthen urban identity [25]. This aligns with research advocating for customized designs over mass-produced solutions, ensuring that urban furniture complements the historical character of its surroundings [27]. Gjuroski [9] and Gravagnuolo and Girard [20] emphasize that street furniture should reinforce the identity of urban landscapes, ensuring that new interventions do not disrupt a place’s historic character. Similarly, our findings indicate that AI-generated classical furniture is perceived as more historically compatible than existing furniture. Köksaldı and Turkan [10] observed that urban furniture in Nicosia’s historical squares often fails to align with cultural texture, which supports our critique of modern urban furniture’s limited contextual sensitivity. This result supports the notion that context-sensitive designs enhance the overall perception of a site’s historical authenticity. Furthermore, Soffritti et al. [11] highlight the role of traditional materials such as cast iron in maintaining historical coherence. This aligns with our study, where wood-metal classical furniture elements were preferred over wood-stone/concrete alternatives, particularly in lighting and benches. This suggests that material selection plays a significant role in expert perceptions of historical compatibility. Another key similarity is the relationship between historical compatibility and user perception. Paiva [38] and Buyukkilic Koşun and Hamamcioglu Turan [22] argue that urban furniture must be visually cohesive with historic settings and evoke a sense of place and emotional connection for users. This study reinforces this perspective, as expert evaluations favored furniture that balanced historical aesthetics with functional usability. The higher preference for AI-generated classical lighting elements, particularly ID 8 (71.0%), aligns with these findings, suggesting that lighting plays a crucial role in defining the atmosphere and identity of historical sites.

4.3. Differences in Findings and Interpretations

Despite these alignments, some differences emerge in the evaluation criteria and prioritization of design elements. Chen [28] and Tiboni et al. [21] emphasize that urban design in historical areas should integrate modern technology while maintaining historical integrity. However, our findings indicate that experts preferred classical over modern designs in historical settings, suggesting a more conservative approach to historical furniture integration. While modern urban furniture was rated highly in general, its preference was notably lower for historical compatibility, contrasting with studies advocating for a hybrid approach incorporating contemporary elements. Additionally, Sipahi and Sipahi [17] focus on sustainability and material selection in urban furniture, advocating for eco-friendly materials such as biodegradable composites. However, our study evaluated historical compatibility primarily regarding aesthetics and material coherence rather than sustainability. The preference for wood–metal over wood–stone/concrete materials in classical urban furniture suggests that experts prioritize traditional aesthetics over sustainability considerations when assessing historical compatibility. Another contrast is evaluating trash cans, where existing furniture was rated lower in compatibility, while AI-generated classical designs were favored. Kou et al. [18] emphasize the role of functional integration in historic districts, arguing that practical elements such as trash cans and benches must be adapted to modern needs while preserving their historical essence. Our findings suggest that while experts recognized historical integration as essential, functionality was a secondary concern in their evaluation of urban furniture, unlike the findings of Kou et al. [18]. Finally, Ye et al. [31] emphasize that AI in urban design may amplify systemic bias without ethical and justice-oriented frameworks. Additionally, Chan [32] provides an essential ethical lens, emphasizing that adopting AI in urban environments must be aligned with urban ethics, co-production of knowledge, and equitable participation. This study supports our study’s underlying methodology, which integrates expert human feedback to ensure culturally grounded and ethically informed AI usage in historical settings.

4.4. Limitations and Future Work

This study offers valuable insights into evaluating urban furniture within historical environments; however, certain limitations must be acknowledged to ensure future research aligns more comprehensively with sustainability principles. One key limitation relates to the composition of the expert group. The participating experts were directly responsible for designing, preserving, and managing historical urban areas—ensuring a high level of contextual knowledge and professional relevance. Moreover, statistical analyses demonstrated a high degree of consistency and agreement among participants, as detailed in the findings section, supporting the internal validity of the results. Still, broader participation from professionals such as environmental psychologists, conservationists, urban designers, and—most importantly—end-users and local communities could significantly enrich future evaluations. Including different stakeholder groups would allow for more nuanced assessments of emotional, functional, and cultural compatibility in urban furniture. Therefore, future studies should consider expanding the pool of participants to include experts such as historians, conservation specialists, urban planners, and a diverse range of user groups and decision-makers. This would enable a more comprehensive and inclusive evaluation of urban furniture design, ensuring that professional expertise and user experience are adequately represented when assessing adaptability and contextual harmony. Moreover, the regional focus of this study, limited to national university settings, may affect the transferability of the findings to other historical urban contexts with different architectural, cultural, and socio-economic conditions. Expanding the geographic scope to include comparative case studies from various regions can help explore how diverse cultural and environmental factors influence sustainable urban design. While this study focused on aesthetic and material compatibility, future research should integrate participatory frameworks that evaluate emotional perception, user experience, and inclusive design metrics across various demographic profiles. Exploring eco-friendly materials, smart urban furniture solutions, and adaptive reuse strategies could provide new perspectives on maintaining historical authenticity while addressing modern urban needs.

While AI-assisted design tools offer significant opportunities in urban planning processes, it is also essential to recognize that these systems may be vulnerable to security threats. Recent studies have shown that AI systems can be manipulated through methods such as data poisoning and adversarial attacks, which may compromise the reliability of design outputs. In the context of smart city applications, such attacks on AI-based urban furniture design systems could lead to the implementation of inappropriate or unsafe designs [51]. Moreover, if AI-integrated Internet of Things (IoT) infrastructures are targeted, broader public safety and sustainability risks may emerge. Therefore, integrating robust security measures into AI-driven design processes has become increasingly critical for future research and practical applications.

Finally, AI-generated urban furniture alternatives can serve as practical preliminary tools for municipalities and urban designers, particularly in historically sensitive contexts. These designs facilitate rapid visualization of stylistic and material options compatible with cultural heritage, offering valuable input in early-stage decision-making processes. However, practical constraints—such as material costs, fabrication challenges, and regulatory requirements in protected areas—may limit their direct implementation. Therefore, AI-generated outputs should be regarded as supportive design references that require contextual adaptation by professionals. Future work should also investigate how diverse stakeholder feedback—from residents, tourists, and accessibility-focused users—can be integrated into AI generation processes to improve inclusivity and social acceptance. Exploring the long-term impact of AI-assisted urban furniture designs in historical areas—including user experiences, public perception, and durability over time—could offer deeper insights into how these designs interact with historical sites in real-world applications. By addressing these limitations, future research can contribute to a more holistic understanding of historical urban furniture design, ensuring that heritage conservation and contemporary urban development remain in harmonious balance.

5. Conclusions

This study provides a comparative analysis of existing and AI-generated urban furniture (modern and classical styles) within the historical setting of Harput Sarahatun Mosque Square in Elazığ, Türkiye, evaluated by expert architects and landscape architects. The research aims to understand how urban furniture can align with historical continuity and sustainable design principles, addressing aesthetic and functional concerns and cultural and environmental sustainability.

The evaluation of existing urban furniture primarily focuses on historical compatibility and material selection, with experts emphasizing aesthetic appeal, comfort, and simplicity rather than historical integration. Findings reveal significant differences in expert evaluations across the three categories of urban furniture—existing, modern, and classical-style designs. Existing urban furniture is primarily evaluated through its practicality and material durability, reflecting a utilitarian approach that often overlooks cultural and historical integration. Experts noted that while these pieces offer comfort and simplicity, they lack sensitivity to the historical and cultural context, thus limiting their contribution to place-based sustainability. Modern AI-generated urban furniture emphasizes design innovation, minimalism, and material simplicity. Experts perceived these designs as structurally clear and universally appealing but essentially detached from local identity or cultural heritage. While they represent advances in material efficiency and functional ergonomics, their limited cultural referencing raises questions about their long-term social sustainability in heritage environments. In contrast, AI-generated classical-style furniture scored highest in historical compatibility, cultural coherence, and aesthetic harmony evaluations. These designs exhibited more substantial alignment with the historical texture of the site, with ornamentation, proportions, and material choices contributing to a more profound sense of continuity. The higher total explained variance in expert evaluations of classical designs suggests they elicit more consistent and structured perceptions, reinforcing their potential to support cultural and historical sustainability.

Overall, the results underscore the importance of a context-sensitive approach to sustainable urban design. Future urban furniture should strive to integrate historical aesthetics with adaptive, environmentally responsible materials, supporting both functional needs and cultural identity. By doing so, urban spaces can promote environmental sustainability, cultural resilience, and social cohesion—core principles of sustainable development. This research contributes to the discourse on sustainable urban design by highlighting the potential of AI-assisted design tools to bridge heritage conservation and innovation in public space planning.