Evaluation of Emotional Vitality Characteristics in Urban Commercial Complexes Based on Multi-Criteria Decision-Making Method: A Case Study of Five Urban Complexes in Beijing

Abstract

Aiming at a critical gap in evaluating human-centered spatial quality during urban stock renewal, this study aimed to develop a quantitative methodology to evaluate emotional vitality in urban commercial complexes. Focusing on five representative Beijing complexes selected according to Beijing urban renewal best practices, the research establishes an integrated evaluation framework combining emotional attachment theory with multi-criteria decision-making method. The methodology employs scales to measure emotional attachment, and implements an improved TOPSIS model integrating Entropy Weight and Grey Relational Analysis of Multi-Criteria Decision-Making method for comprehensive assessment. As a result, key findings demonstrate that small-scale built environment features significantly enhance spatial vitality through emotional attachment mechanisms. Material characteristics, including historical material reinterpretation and innovative structural elements, prove fundamental in fostering attachment, while non-material features exhibit dynamic influences that evolve with temporal and contextual factors. The study reveals that successful emotional attachment requires balanced integration of physical and social features, with material characteristics serving as the foundation for sustained emotional vitality. The research contributes an evidence-based evaluation system that bridges theoretical constructs from environmental psychology with practical urban design applications. By objectively quantifying emotional attachment and identifying specific spatial features that enhance these emotional experience, this methodology provides valuable tools for urban designers, planners, and governors seeking to create more engaging and humanized commercial environments. The framework further offers scientific support for decision-making in urban renewal projects and establishes a replicable approach for vibrant urban space evaluation.

1. Introduction

In the early 21st century, the speed and scale of China’s urbanization once positioned it alongside the rapid advancement of U.S. high technology as two major factors influencing the development of human society. In recent years, Chinese cities have gradually transitioned to a stage of development focused on stock renewal. Under the policy guidance of connotative growth [1,2], improving the quality of life for residents has become a key focus in the construction of first-tier and new first-tier cities [3,4,5], such as Beijing, Shanghai and Guangzhou [6]. Beijing’s 2021–2025 Territorial Spatial Planning policy has emphasized a reduction in quantity and improvement in quality renewal strategy. It highlights the value of small-scale and decentralized spaces such as urban complexes in fostering urban vitality. Transforming abandoned built environments into diverse urban complexes has proven to be a successful approach in creating small-scale vibrant urban spaces [7,8], gaining public recognition and becoming important emotional anchor points in urban life [9]. From 2022 to 2024, in public-participated votes for Beijing’s best urban renewal practices, aside from livelihood projects such as houses for low income people, the majority were urban commercial complexes that created vibrant spaces through innovative functions. This further emphasizes the need for research and evaluation of methods to enhance vitality in these renewal complexes.

This study selects five representative complexes from Beijing’s outstanding urban renewal practices as research areas to validate the effectiveness of the aforementioned evaluation method, including 751D·PARK, THE BOX, Galleria-Instreet, SHOKAI LONG Street, and Beijing Fun. First, based on emotional attachment theory, the study constructs a scale targeting the spatial characteristics of the built environment of these complexes to measure the degree, dimensions, and intensity of people’s attachment to their spatial features. SPSS 26.0 software is then used to process the scale data for a vitality evaluation index for attachment. An improved TOPSIS method (Technique for Order Preference by Similarity to Ideal Solution), combined with EW (Entropy Weight) and GRA (Grey Relational Analysis), is selected from MCDM methods to evaluate the vitality levels of the five complexes. Finally, the correlation between the evaluation ranking results and the spatial features of the evaluation indicators is analyzed to further explore the commonalities and differences in spatial features that promote attachment across the five vitality complexes.

The study aims to make theoretical, methodological, and practical promotion to urban vitality research by incorporating emotional attachment between people and commercial built environments into the existing frameworks, developing an evaluation framework for urban commercial complexes using scales and an integrated EW–GRA–TOPSIS MCDM model to offer emotional oriented guidance for future vitality space design and evaluation.

2. Literature Review

2.1. Emotion as Guiding Value: Trends and Demands in Urban Vitality Research

Vibrant spaces have always been associated with density of people, diverse activities, and integrated functions [10,11,12] in interdisciplinary research for decades. This makes urban commercial complexes suitable objects for studying urban vitality [13]. Although no consensus on the definition of “urban vitality” has been reached, the term “vitality” implies that such spaces should aiming at creating lively built environments and promoting genuine human experiences. Existing theoretical and practical research on urban vitality across multiple disciplines consistently demonstrates the significance of emotion in the promotion and development of a vibrant space. For example, Jacobs, Gehl and Alexander have argued that genuine vitality actually transcends physical functionality [11,12,14,15,16,17]. It’s about creating spaces where people form emotional bonds with their surroundings. While these urban researchers saw vitality emerging from spaces that attract diverse activities, researchers from geography defined it as the extent to which a place makes people feel alive [18]. Meanwhile, some interdisciplinary researchers working on cognitive architecture further focus on revealing the specific pathways and mechanisms through which vibrant spaces affect human emotions. Their work demonstrates how environmental features unconsciously shape our emotional responses [19]. This theoretical evolution reveals that vibrant spaces grow through emotional attachment mechanisms.

On the other hand, research methods tell a different story. Relevant practices have been preoccupied with two-dimensional metrics such as mobile data [20,21], heat maps [22,23], social media footprints [24,25]. As a result, they reduce complex human experiences to flat indicators [26,27,28]. Though these digital methods offer certain scalability, they inevitably miss what makes spaces truly attach to people. The emotional connections that the above researchers prioritize get lost. This gap becomes especially apparent when evaluating urban complexes. Existing research rarely captures how people actually experience three-dimensional spaces in their daily lives. As a result, we either get qualitative descriptions lacking scientific rigor or abstract data devoid of spatial meaning. What’s missing is precisely what matters most, to understand the physical attributes that trigger genuine emotional experiences [4,9,29].

In a word, in order to create truly vibrant urban spaces, methods that can capture how people emotionally attached to their environment at human scale are in urgent need. This requires moving beyond flat representations to embrace interdisciplinary approaches that honor the full dimensionality of human experience.

2.2. Multidisciplinary Support for Indicator Quantification: Definition and Measurement of Emotional Attachment

Based on the above researchers, the growing need to quantify emotional responses in vitality researches has led researchers to draw on interdisciplinary theories and scales of emotional attachment [30]. These approaches take emotion as the core within a multidimensional framework involving perception, cognition, and affective bonds. While built environment fields like architecture and urban planning have explored attachment through behavioral studies and phenomenological approaches [19,31,32,33,34], fields such as psychology and human geography have contributed systematic frameworks and validated measurement scales [35,36]. As the “place attachment” framework and its associated scales have been particularly valuable for quantifying people-space connections [37,38,39], it provides this study with valuable logical path.

Thus, to reliably construct indicators for evaluating emotional attachment in vibrant urban spaces, this study employs a combination of established and developed scales. The Place Attachment Scale helps capture the intensity of people-space bonds [40,41], while the PANAS scale measures the positive and negative dimension of emotional experiences [42,43]. Additionally, a specific scale was developed and used to evaluate attachment to specific spatial features, which incorporating spatial features like materials and forms, socio-cultural qualities, and contemporary digital characteristics like changeability. Such scales have been successfully applied in studies of architecture, urban, and landscape environments both domestically and internationally [39,44,45,46,47]. The series scales have undergone expert assessment and reliability testing [39,44,45]. The results provide the item pool for the evaluation indicators in the following MCDM model.

In this way, the study establishes a methodological foundation for quantifying emotional attachment. This multidisciplinary perspective not only addresses the previously overlooked emotional dimension in vitality research but also generates quantifiable indicators for subsequent evaluation models, which enables a more nuanced understanding of how people form attachment to vibrant urban spaces.

2.3. Logical Pathways for Evaluation: Multi-Criteria Decision-Making Methods in Urban Studies

To evaluate how the above emotional indicators influence vitality and enable comparative analysis across spaces, multi-criteria decision-making (MCDM) methods offer valuable logical and analytical support [48,49]. Though approaches like AHP and FAHP are commonly used in space evaluation [50,51], this study adopts an integrated EW-GRA-TOPSIS model that better addresses the complexity of emotion as well as urban systems [52,53].

The TOPSIS method ranks alternatives by measuring relative proximity of each alternative to ideal solutions. When combining entropy weighting and grey relation analysis, it creates a more robust framework that helps maximize data utility while minimize subjective bias. Its straightforward implementation and adaptability have made TOPSIS particularly valuable for evaluating complex urban space related problems [54,55,56]. It has been successfully used in the study of urban sustainability, resilience, and vitality studies [54,55,57].

Recent advances have further integrated improved TOPSIS with GIS and machine learning techniques, expanding its capacity to handle multidimensional urban data [58,59,60]. Besides, researchers have begun applying this methodology to architectural environments and people-space interactions at smaller district levels [55,61,62].

Despite these developments, current research rarely employs TOPSIS to examine how specific spatial features affect emotional experiences, which turns out as a critical dimension of urban vitality that this study specifically addresses. This gap represents an important opportunity to advance our understanding of the emotional value of vibrant urban spaces.

3. Materials and Methods

3.1. Study Areas

Based on the research objectives, this study selected five representative vibrant urban commercial complexes from Beijing’s 2023 urban regeneration best practices as research subjects: 751D·PARK, THE BOX, Galleria-Instreet, SHOKAI LONG Street, and Beijing Fun. They represent relatively typical cases of contemporary commercial complexes. Urban commercial complexes, as spatial types that meet the elements of creating such vitality spaces, have become exemplary cases for practical implementation. Emerging urban commercial complex renewal practices have moved beyond traditional multi-functional approaches and begun exploring how to enhance vitality of renewed built environments by evoking multi-dimensional perceptions and meeting people’s advanced emotional needs. Specifically, the names and their capitalization of the five complexes used in the study align with their official English designations. These five complexes were all regenerated from existing structures into distinctive commercial spaces with unique characteristics. THE BOX and Galleria-Instreet were transformed from underutilized commercial spaces; Beijing Fun originated as a traditional commercial street in a historic district; SHOKAI LONG Street was repurposed from vacant, low-efficiency apartment buildings; and 751D·PARK was regenerated from a former industrial factory site.

Although differing in their original conditions and regeneration processes, all cases emphasize human-scale spatial design and have begun exploring how the built environment can respond to people’s increasingly complex emotional needs to enhance spatial vitality. Among them, THE BOX and 751D·PARK specifically target younger demographics in their regeneration strategies, attempting to stimulate vitality through youth-oriented approaches while comprehensively enhancing the richness and diversity of both indoor and outdoor spaces. Galleria-Instreet and SHOKAI LONG Street, focusing on diverse community experiences, prioritize users’ communal public life and activities to create vibrant neighborhood spaces. Beijing Fun integrates multiple regeneration objectives—serving as a new urban cultural landmark, a Chinese-style lifestyle experience zone, and an international consumption center—evolving into a distinctive space rooted in Beijing’s local character while possessing international appeal. Locations of each complex are shown in Figure 1.

3.2. Evaluation Methodology

3.2.1. Development of Evaluation Indicators

To establish the evaluation indicators for complexes (

Table 1. Evaluation Indicators and Corresponding Measurement Methods.

Evaluation Indicators		Corresponding Measurement Methods
Degree	Place attachment	Place Attachment Scale(PA Scale)	(1 point indicates strongest disagreement with the statement, 7 points indicate strongest agreement with the statement)
			Everything about this place is a reflection of me.		This place reflects the type of person I am.
			This place says very little about who I am.		As far as I am concerned, there are better places to be than in this place.
			I feel relaxed when I’m in this place.		The spiritual nature of the area ties me to this place.
			I feel happiest when I’m in this place.		I feel that this place is my home.
			This place is my favorite place to be.		I intend to continue staying in or around this place for the next few years.
			I really miss this place when I’m away from it for too long.		I have the feeling that this place constitutes a security base for me.
			I feel that I can really be myself in this place.		I feel a connection to the visual landscape of the area.
			This place is the best place for doing the things I enjoy most.		This place is an important part of my life.
			For doing the things that I enjoy most, no other place can compare to this place.		I feel proud of this place.
			This place is not a good place to do the things I most like to do.		I am totally involved and committed to my school, classmates and neighborhood.
Dimension	Positive	Positive and Negative Affect Scale(PANAS)	(Comprising 10 positive emotion and 10 negative emotion adjectives, measured using a 5-point Likert scale where 1 point indicates almost no such emotional experience, and 5 points indicate intense such emotional experience)
	Negative		Positive adjectives: interested, excited, strong, enthusiastic, proud, inspired, determined, active, alert, attentive
			Negative adjectives: distressed, upset, guilty, scared, hostile, irritable, ashamed, nervous, jittery, afraid
Intensity of attachment to specific vibrant environment features	Environment Feature Emotional Attachment Intensity Scale		(1 point indicates almost no attachment derived from the feature, 7 points indicate strong attachment derived from the feature)
			Material	Color	Natural Elements	Form and Structure
			Privacy	Diversity	Sociability	Territoriality
			Playability	Uniqueness	Changeability

), this study integrates the place attachment scale, the PANAS scale, and the scale of attachment intensity for specific built environment features to investigate the degree, dimension, and intensity of people’s attachment to specific built environment features across five urban renewal complexes.

Regarding the indicator for the degree of attachment, the study employs the Place Attachment Scale developed by Scannell and colleagues for measurement and extraction [40,63]. The theoretical framework of the scale provides a comprehensive description of the interactive process between people, the built environment, and emotions, aligning with the logic of built environment research. Based on the sample of this study, the reliability coefficient (Cronbach’s α) of the scale was 0.94, indicating excellent internal consistency.

For the indicator concerning the valence, the measurement is based on the PANAS scale, aimed at identifying the positive or negative valence of individuals’ overall emotional response towards different complexes. In this study, the Positive Affect scale demonstrated a Cronbach’s α of 0.89, while the Negative Affect scale showed a reliability coefficient of 0.84, both indicating good to excellent internal consistency.

For measuring the intensity elicited by specific built environment features, a semi-structured scale was constructed for measurement and extraction. This was based on psychometric scale development pathways, preliminary studies on the environmental features of the five complexes, and a Likert-type scoring method. This scale has previously been validated in prior related research through expert scoring and assessments of its reliability and validity, and its reliability coefficient was 0.95 in this sample.

The aforementioned scales have been validated in numerous built environment studies both domestically and internationally, demonstrating their effectiveness in revealing the characteristics of attachment within spaces. The collected scale data were processed and analyzed using SPSS 26.0, including descriptive statistical analysis and correlation analysis. For the former, means and standard deviations were calculated for the indicators of degree, dimension, and specific features intensity. This serves, on one hand, to describe the overall characteristics of the subjects’ emotional attachment to the five complexes and, on the other hand, to provide sorting indicators for the subsequent evaluation model. For the latter, correlation analysis was employed to explore the degree of attachment between the specific built environment features of the five complexes and attachment indicators, which provided a systematic basis for subsequent analysis while simultaneously enabling deeper exploration of the interdependences among evaluation factors.

Theoretically, the formation of emotional attachment emerges through coordinated interactions among affective responses and environmental characteristics, where artificial and interactive elements—particularly through the synergy of material, color, and form—serve as primary drivers of positive affect and attachment. Natural elements, while showing limited direct emotional impact, function as essential moderators by enhancing spatial comfort and territorial identity. The network of diversity, playability, and uniqueness sustains engagement through varied experiences, while sociability and territoriality operate complementarily to balance social interaction with personal space needs. This interdependent framework demonstrates that emotional attachment arises not from isolated elements but from the integrated effects of environmental attributes working in concert.

3.2.2. Construction of Evaluation Model

It should be acknowledged that the evaluation outcomes are intrinsically tied to the developed scales and their corresponding assessment indicators. Although the reliability and validity of these scales have been established in previous research, certain limitations and uncertainties remain in the evaluation process. These stem from both the inherent subjectivity of researchers and designers, as well as objective factors such as variations in scale data collection. Such uncertainties may impact the accuracy of assessing attachment within urban spatial dimensions.

To mitigate these limitations and enhance evaluation robustness, this study employs the Entropy Weight Method (EW) to objectively weight the aforementioned indicators based on their measured values. Furthermore, we integrate the classic TOPSIS evaluation model with Grey Relation Analysis (GRA), thereby addressing the singular evaluation logic of traditional TOPSIS. This integrated approach improves the objectivity of assessing vitality creation characteristics in urban renewal complexes. The operational procedures and computational methodology of this enhanced evaluation model are detailed below.

• Weighting of Evaluation Indicators with the Entropy Weight Method (EW)

Existing research demonstrates that employing the Entropy Weight Method (EW) rather than the Delphi method or Analytic Hierarchy Process (AHP) to determine evaluation indicator weights effectively reduces the subjectivity inherent in the traditional TOPSIS model. This approach enhances the objectivity of evaluations for systems comprising interconnected indicators. The EW-based methodology has been successfully applied by numerous scholars in urban spatial assessment and related fields, demonstrating its capacity to minimize subjective bias in evaluation processes. Following this rationale, the procedure for determining weights of the aforementioned evaluation indicators using the Entropy Weight Method in this study is outlined as follows:

This study involves $m$ complexes to be evaluated, each characterized by $n$ built-environment-related evaluation indicators. The value of the $j$-th indicator for the $i$-th complex is denoted as $c_{ij}\left(1\le i\le m,1\le j\le n\right)$. The matrix $C$ is constructed as follows:

$$C=\left(\begin{array}{ccc}{c}_{11}& \dots & c_{1n}\\ ⋮& \ddots & ⋮\\ c_{m1}& \dots & c_{mn}\end{array}\right)$$

(1)

where $c_{ij},1\le i\le m,1\le j\le n,m=5,n=14$.

The judgment matrix $C={\left(c_{ij}\right)}_{m\times n}$ is standardized according to the 0–1 transformation, and the standardized matrix $C^{’}={\left(c_{ij}^{’}\right)}_{m\times n}$.

$$C^{’}=\left(\begin{array}{ccc}{c}_{11}^{’}& \dots & c_{1n}^{’}\\ ⋮& \ddots & ⋮\\ c_{m1}^{’}& \dots & c_{mn}^{’}\end{array}\right)$$

(2)

This study classifies negative affect as a negative indicator and all other measures as positive indicators. We therefore preprocess the positive indicators with Formula (3) and the negative indicator with Formula (4).

$$c_{ij}^{’}={\displaystyle \frac{c_{ij}-\underset{i}{\min } c_{ij}}{\underset{i}{\max } c_{ij}-\underset{i}{\min } c_{ij}}}\underset{i}{\max } c_{ij}\ne \underset{i}{\min } c_{ij}{c}_{ij}^{’}=1\underset{i}{\max } c_{ij}=\underset{i}{\min } c_{ij}$$

(3)

$$c_{ij}^{’}={\displaystyle \frac{\underset{i}{\max } c_{ij}-c_{ij}}{\underset{i}{\max } c_{ij}-\underset{i}{\min } c_{ij}}}\underset{i}{\max } c_{ij}\ne \underset{i}{\min } c_{ij}{c}_{ij}^{’}=1\underset{i}{\max } c_{ij}=\underset{i}{\min } c_{ij}$$

(4)

Calculate the proportion $q_{ij}$ that the $i$-th complex occupies in the $j$-th indicator. The formula is as follows

$$q_{ij}={\displaystyle \frac{c_{ij}^{’}}{{\displaystyle \sum _{i=1}^m{c}_{ij}^{’}}}},i=1,2,\dots ,m; j=1,2,\dots ,n$$

(5)

Calculate the entropy value $E_j$ of each indicator, the formula is as follows:

$$E_j=-k{\displaystyle \sum _{i=1}^m{q}_{ij}\ln q_{ij}}, j=1,2,\dots ,n$$

(6)

where $k={\displaystyle \frac{1}{\ln m}},0\le E_j\le 1$.

The weight for each indicator is subsequently calculated. Following this, the information entropy redundancy $d_j$ for an indicator is defined by the formula:

$$d_j=1-E_j,j=1,2,\dots ,n$$

(7)

Using the information entropy redundancy $d_j$, the weight ${\omega }_j$ of the $j$-th indicator is calculated as follows:

$${\omega }_j={\displaystyle \frac{d_j}{{\displaystyle \sum _{j=1}^n{d}_j}}},j=1,2,\dots ,n$$

(8)

Following the aforementioned steps, the respective weights of the 14 evaluation indicators—derived from the scales to describe the degree, dimension, and intensity of spatial emotional attachment—were calculated. These weights will be used for subsequent ranking computations in the GRA-TOPSIS model.

• Improving TOPSIS model by incorporating the Grey Relational Analysis (GRA).

The quantitative assessment of emotional attachment constitutes a multi-attribute decision-making problem. This study enhances the TOPSIS method with Grey Relation Analysis to better capture non-linear relationships among interdependent indicators. The integrated approach follows three steps: identifying ideal solutions via TOPSIS, calculating relational degrees with GRA, and ranking complexes by attachment strength. This methodology provides a more nuanced evaluation of urban spatial attachment. The specific computational steps are as follows:

The judgment matrix $C={\left(c_{ij}\right)}_{m\times n}$ is standardized to yield a normalized matrix $X={\left(x_{ij}\right)}_{m\times n}$, whose elements are computed as follows:

$$x_{ij}={\displaystyle \frac{c_{ij}^{\prime }}{\sqrt{\sum _{i=1}^m{c_{ij}^{’}}^2}}},1\le i\le m,1\le j\le n$$

(9)

The weight vector ${\omega }_j=\left({\omega }_1,{\omega }_2\dots {\omega }_n\right)$ obtained in Equation (8) is multiplied by the standardization matrix $X={\left(x_{ij}\right)}_{m\times n}$ to get a weighted standardization matrix as follows:

$$Y=\left(\begin{array}{cccc}{\omega }_1{x}_{11}& {\omega }_2{x}_{12}& \dots & {\omega }_n{x}_{1n}\\ {\omega }_1{x}_{21}& {\omega }_2{x}_{22}& \dots & {\omega }_n{x}_{2n}\\ ⋮& ⋮& \dots & ⋮\\ {\omega }_1{x}_{m1}& {\omega }_2{x}_{m2}& \dots & {\omega }_n{x}_{mn}\end{array}\right)=\left(\begin{array}{cccc}{y}_{11}& y_{12}& \dots & y_{1n}\\ y_{21}& y_{22}& \dots & y_{2n}\\ ⋮& ⋮& \dots & ⋮\\ y_{m1}& y_{m2}& \dots & y_{mn}\end{array}\right)$$

(10)

The positive ideal solution $y^{+}$ and negative ideal solution $y^{-}$ is calculated as follows, where $J^{+}$ is the set of positive indicators while $J^{+}$ is the set of negative indicators:

$$y^{+}=\left[\left(\underset{1\le i\le m}{\max }{y}_{ij}|j\in J^{+}\right),\left(\underset{1\le i\le m}{\min }{y}_{ij}|j\in J^{-}\right)\right]=\left(y_1^{+},y_2^{+},\dots ,y_n^{+}\right)$$

(11)

$$y^{-}=\left[\left(\underset{1\le i\le m}{\min }{y}_{ij}|j\in J^{+}\right),\left(\underset{1\le i\le m}{\max }{y}_{ij}|j\in J^{-}\right)\right]=\left(y_1^{+},y_2^{+},\dots ,y_n^{+}\right)$$

(12)

In this study, the positive ideal solution $y^{+}$ corresponds to a hypothetical urban renewal complex that exhibits the optimal capacity for fostering emotional attachment and achieves the highest possible level of such attachment. Conversely, the negative ideal solution $y^{-}$ represents the theoretical construct with the poorest capacity and lowest level of attachment.

The Euclidean distances $r^{+}$ and $r^{-}$, measuring the separation of each complex from the positive and negative ideal solutions respectively, are calculated. A smaller value of $r^{+}$ indicates that a complex is closer to the positive ideal solution and is associated with a higher intensity of attachment. Conversely, a smaller value of $r^{-}$ signifies proximity to the negative ideal solution and corresponds to a lower attachment intensity.

$$r_i^{+}=\sqrt{{\displaystyle \sum _{j=1}^n{\left[\left(y_{ij}-y_j^{+}\right)\right]}^2}},1\le i\le m,1\le j\le n$$

(13)

$$r_i^{-}=\sqrt{{\displaystyle \sum _{j=1}^n{\left[\left(y_{ij}-y_j^{-}\right)\right]}^2}},1\le i\le m,1\le j\le n$$

(14)

Calculation of the grey relation degree:

$$d_{ij}^{+}={\displaystyle \frac{\underset{i}{\min } \underset{j}{\min }|y^{+}-y_{ij}|+\rho \underset{i}{\max } \underset{j}{\max }|y^{+}-y_{ij}|}{|y^{+}-y_{ij}|+\rho \underset{i}{\max } \underset{j}{\max }|y^{+}-y_{ij}|}}$$

(15)

$$d_{ij}^{-}={\displaystyle \frac{\underset{i}{\min } \underset{j}{\min }|y^{-}-y_{ij}|+\rho \underset{i}{\max } \underset{j}{\max }|y^{-}-y_{ij}|}{|y^{-}-y_{ij}|+\rho \underset{i}{\max } \underset{j}{\max }|y^{-}-y_{ij}|}}$$

(16)

$$d_i^{+}={\displaystyle \frac{1}{n}}{\displaystyle \sum _{j=1}^n{d}_{ij}^{+}},1\le i\le m$$

(17)

$$d_i^{-}={\displaystyle \frac{1}{n}}{\displaystyle \sum _{j=1}^n{d}_{ij}^{-}},1\le i\le m$$

(18)

Dimensionless processing weighted the distances $r^{+}$ and $r^{-}$, and the grey correlation $d_i^{+}$ and $d_i^{-}$.

$$R_i^{+}={\displaystyle \frac{\max r_i^{+}-r_i^{+}}{\max r_i^{+}-\min r_i^{+}}},1\le i\le m$$

(19)

$$R_i^{-}={\displaystyle \frac{\max r_i^{-}-r_i^{-}}{\max r_i^{-}-\min r_i^{-}}},1\le i\le m$$

(20)

$$D_i^{+}={\displaystyle \frac{d_i^{+}-\min d_i^{+}}{\max d_i^{+}-\min d_i^{+}}},1\le i\le m$$

(21)

$$D_i^{-}={\displaystyle \frac{d_i^{-}-\min d_i^{-}}{\max d_i^{-}-\min d_i^{-}}},1\le i\le m$$

(22)

Integrate the above dimensionless distance and dimensionless grey correlation degree results:

$$S_i^{+}=\alpha R_i^{+}+\beta D_i^{+},1\le i\le m$$

(23)

$$S_i^{-}=\alpha R_i^{-}+\beta D_i^{-},1\le i\le m$$

(24)

$$\alpha +\beta =1,\alpha >0,\beta >0$$

Calculation of the relative closeness $A_i^{+}$ of the complex

$$A_i^{+}={\displaystyle \frac{S_i^{+}}{S_i^{+}+S_i^{-}}},1\le i\le m$$

(25)

The higher the $A_i^{+}$, the closer the complex $C_i$ is to the positive ideal solution, indicating that its ability to promote attachment is better, and the intensity is also higher.

A systematic sensitivity analysis confirms the robustness of the EW-GRA-TOPSIS framework, demonstrating minimal rank volatility under parameter variations. Testing ±10% perturbations in entropy-derived weights resulted in negligible ranking changes (<±2 positions), validating the stability of objective weighting. The GRA resolution coefficient (ζ) exhibited optimal stability within 0.5–0.7 range, with extreme values causing predictable compression effects without altering fundamental rankings. The hybrid model demonstrates superior resilience to parameter fluctuations compared to individual methodologies, ensuring reliable comparative assessment across urban complexes.

3.3. Data Collection

To ensure a rigorous quantitative assessment, the research team developed a comprehensive data collection protocol based on established emotional attachment theory. The survey instrument was refined through three rounds of pre-testing involving 45 participants across different urban complexes.

The sampling strategy employed a stratified random approach during April–May 2024, with data collection scheduled across two distinct phases: 15–21 April and 10–16 May 2024. This temporal distribution ensured representation of both regular weekdays and weekends at each site. Survey locations were systematically selected to cover all functional zones within the five complexes, including retail areas, public plazas, transitional spaces, and entertainment zones.

The team members conducted the surveys during four daily time slots (9–11 AM, 11 AM-1 PM, 2–4 PM, 5–7 PM) using a randomized intercept technique. The recruitment protocol specified that every fifth adult passing through designated survey zones would be invited to participate, with gender and age quotas monitored in real time to ensure demographic diversity.

The data validation process implemented multiple quality checks: incomplete responses (>10% missing data), straight-line patterns, inconsistent responses to reverse-coded items, and completion times under 3 min (determined through pilot testing) were excluded. The final valid sample sizes reflect this rigorous screening: 751D·PARK (n = 103), THE BOX (n = 118), Galleria-Instreet (n = 101), SHOKAI LONG Street (n = 94), and Beijing Fun (n = 108), achieving a combined response rate of 78.3% from initially distributed questionnaires.

3.4. Research Framework

In summary, this study employs a multi-method analytical approach to comprehensively evaluate emotional attachment characteristics in the small-scale built environments of urban complexes. The research design follows a systematic pathway of “defining evaluation indicators—establishing evaluation models—conducting evaluative analysis” to assess attachment across five case study sites. The rationale for methodological integration stems from the multidimensional nature of emotional attachment, which requires complementary analytical perspectives to fully capture its complexity.

The integrated methodology combines scale analysis, Entropy Weight Method, and enhanced TOPSIS-GRA modeling with systematically incorporated on-site observational interviews. These interviews were conducted during both pre-survey contextual understanding and post-evaluation results interpretation phases, providing crucial qualitative insights that complemented the quantitative data. This mixed-methods approach ensured the human-centered emotional experiences were preserved throughout the research process while maintaining analytical rigor. The methodological triangulation enhances findings’ robustness by mitigating single-method limitations, delivering both quantitative rankings and qualitative understanding of how specific built environment characteristics generate emotional attachment. The logical framework for the evaluation is illustrated in Figure 2.

4. Results

The calculated results for various attachment evaluation indicators across the five complexes are presented in

Table 2. Results of emotional attachment indicators of the five complexes.

Emotional Attachment Index		751D·PARK		THE BOX		Galleria-Instreet		SHOKAI LONG Street		Beijing Fun
		Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Attachment Degree	Place Attachment	3.99	0.97	3.95	0.98	3.73	1.10	4.18	1.04	4.22	1.13
Attachment Dimension	Positive Affect	2.63	0.76	2.62	0.74	2.36	0.77	2.64	0.89	2.58	0.84
	Negative Affect	1.36	0.43	1.31	0.37	1.29	0.48	1.21	0.31	1.30	0.42
Attachment Intensity to Specific Environment Feature	Material	4.89	1.55	4.81	1.51	4.55	1.53	5.04	1.67	4.88	1.75
	Color	5.37	1.48	5.52	1.45	4.80	1.48	5.39	1.52	5.21	1.73
	Natural Elements	4.88	1.52	4.47	1.83	4.73	1.55	5.01	1.55	5.02	1.73
	Form and Structure	5.38	1.43	5.25	1.41	4.93	1.57	5.33	1.50	5.17	1.74
	Privacy	3.48	1.51	3.78	1.74	3.50	1.63	3.85	1.81	4.46	1.73
	Diversity	5.31	1.46	5.25	1.53	4.50	1.68	4.91	1.64	4.92	1.75
	Sociability	4.79	1.74	5.07	1.51	4.92	1.59	4.74	1.68	4.73	1.72
	Territoriality	4.50	1.75	4.42	1.74	4.28	1.68	4.29	1.80	5.21	1.63
	Playability	4.01	1.72	4.48	1.82	3.94	1.71	4.69	1.80	4.00	1.80
	Uniqueness	5.33	1.73	5.50	1.50	4.30	1.64	4.86	1.81	5.06	1.75
	Changeability	4.71	1.77	5.08	1.63	4.26	1.77	4.63	1.91	4.55	1.75

. The results were further used for evaluation model computation.

Following Equations (1)–(25), the constructed EW-GRA-TOPSIS evaluation model has been conducted through the steps of calculating weight of indicator (

Table 3. Weight of each indicator.

Evaluation Index	${\mathit{E}}_{\mathit{j}}$	${\mathit{\omega }}_{\mathit{j}}$
Place Attachment	0.9831	6.68%
Positive Affect	0.9836	6.49%
Negative Affect	0.9847	6.06%
Material	0.9857	5.66%
Color	0.9848	6.02%
Natural Elements	0.9827	6.83%
Form and Structure	0.9843	6.20%
Privacy	0.9786	8.47%
Diversity	0.9834	6.57%
Sociability	0.9766	9.26%
Territoriality	0.9763	9.38%
Playability	0.9742	10.22%
Uniqueness	0.9842	6.25%
Changeability	0.9850	5.94%

), determining positive ideal solution and negative ideal solution of each indicator (

Table 4. Positive ideal solution and negative ideal solution of each indicator.

Evaluation Index	Positive Ideal Solution	Negative Ideal Solution
Place Attachment	0.134	0.067
Positive Affect	0.130	0.065
Negative Affect	0.121	0.061
Material	0.113	0.057
Color	0.120	0.060
Natural Elements	0.137	0.068
Form and Structure	0.124	0.062
Privacy	0.169	0.085
Diversity	0.131	0.066
Sociability	0.185	0.093
Territoriality	0.188	0.094
Playability	0.204	0.102
Uniqueness	0.125	0.062
Changeability	0.119	0.059

), calculating and normalizing of the distances between each complex’s attachment and the ideal solutions (

Table 5. Distance from the emotional attachment of each complex to the positive and negative ideal solution.

Complex	${\mathit{R}}_{\mathit{i}}^{+}$	${\mathit{R}}_{\mathit{i}}^{-}$
751D·PARK	0.181	0.157
THE BOX	0.139	0.193
Galleria-Instreet	0.247	0.067
SHOKAI LONG Street	0.151	0.197
Beijing Fun	0.154	0.188

), calculating grey relational degree (

Table 6. Grey relational degree of each complex.

Complex	${\mathit{D}}_{\mathit{i}}^{+}$	${\mathit{D}}_{\mathit{i}}^{-}$
751D·PARK	0.669	0.636
THE BOX	0.734	0.573
Galleria-Instreet	0.459	0.906
SHOKAI LONG Street	0.756	0.574
Beijing Fun	0.712	0.591

), and determining relative closeness to show final rank results (

Table 7. Relative closeness and the vitality evaluation rank of five complexes.

Complex	${\mathit{S}}_{\mathit{i}}^{+}$	${\mathit{S}}_{\mathit{i}}^{-}$	${\mathit{A}}_{\mathit{i}}^{+}$	Rank
751D·PARK	1.659	1.248	0.571	4
THE BOX	1.963	1.015	0.659	2
Galleria-Instreet	1.000	2.000	0.333	5
SHOKAI LONG Street	1.944	1.002	0.660	1
Beijing Fun	1.856	1.062	0.636	3

).

The results indicate that SHOKAI LONG Street ranked first among the five complexes in the evaluation, demonstrating that users have established the highest level of attachment with this urban complex compared to the others. THE BOX ranked second, with score (0.659) showing minimal difference from SHOKAI LONG Street. Galleria-Instreet ranked last, suggesting a generally lower degree of attachment among users to this location. By integrating the correlation analysis results of scales for the five complexes with on-site observations and interviews, the underlying reasons for these rankings can be further analyzed.

5. Discussion

5.1. Overall Characteristics of Emotional Attachment: Analysis Based on Calculated Evaluation Indicators

Combining with the results shown in Table 2, the degree, described by the place attachment indicator, reveals that all five complexes have established strong emotional connections with users (with place attachment values exceeding the median). The standard deviation results reveal variations in attachment levels across complexes, with Beijing Fun showing the greatest diversity and 751D·PARK the most homogeneous user responses. This pattern aligns with existing studies that document similar consistency in affective experiences at street-scale environments, though our findings demonstrate more pronounced individual variations in commercial complexes than typically reported in larger-scale urban research. The positive and negative affect indicators consistently show strong positive emotions and weak negative emotions across all five complexes, with relatively high interpersonal consistency (small standard deviations)—a phenomenon observed in previous studies of successful vibrant spaces but now quantified specifically for commercial complexes.

The emotional attachment intensity toward spatial features, measured through 11 physical and non-physical indicators, generally maintains strong values above median levels while exhibiting considerable individual variations (relatively large standard deviations). This dual pattern of strong overall attachment with significant individual differences echoes findings from streetscape studies, yet reveals greater complexity in commercial complex environments where diverse user groups interact with the same spatial features. The varying attachment intensities across different complexes, when examined alongside evaluation rankings, provide empirical evidence that extends beyond conventional metrics—addressing a recognized gap in small-scale commercial space research where previous studies have primarily focused on behavioral rather than emotional dimensions.

These scale results not only form the foundation for our EW-GRA-TOPSIS evaluation system but also offer validated evidence for the attachment value of built environment features, thereby bridging methodological approaches between established urban vitality research and emerging emotional attachment studies in commercial contexts.

5.2. Ranking of Emotional Attachment and Its Influencing Factors: A Multidimensional Analysis Integrating MCDM Evaluation and Scale Results

The correlation analysis elucidates the contributions of built environment features to both the intensity and dimensions of attachment. By correlating these findings with the aforementioned evaluation rankings, we can further examine the attachment utility of different physical and non-physical characteristics in the built environment.

The correlation analysis across the five mixed-use developments reveals a consistent positive relationship between place attachment and positive emotional experiences, demonstrating that such experiences strengthen individuals’ emotional bonds with the built environment. Furthermore, the findings indicate that a more focused range of emotional dimensions linked to environmental features corresponds to higher overall attachment levels. This pattern is evident in the two top-ranked cases—SHOKAI LONG Street and THE BOX—where only positive emotions showed statistically significant correlations with attachment strength across various environmental attributes, while negative emotions demonstrated no significant relationship. Conversely, lower-ranked complexes exhibited significant inverse correlations between negative emotions and attachment strength across multiple features. Notably, when negative emotional experiences were associated with non-material environmental characteristics, overall attachment levels decreased substantially, as observed in the lowest-ranked case, Galleria-Instreet. Additionally, material features—including materials, color, natural elements, form and structure—consistently contributed to place attachment across all five developments. This effect was particularly pronounced in SHOKAI LONG Street, which achieved the highest attachment rating. These findings highlight the importance of tangible design elements in fostering emotional engagement within small-scale urban revitalization projects.

SHOKAI LONG Street (rank 1) demonstrated particularly strong positive correlations between built environment features and place attachment, with nearly all features showing significant relationships (Figure 3). The material attributes—including materials, color, natural elements, form and structure—exhibited especially pronounced effects, with correlation coefficients all exceeding 0.600. Further analysis revealed significant intercorrelations among these four material attributes (p < 0.01, r > 0.600), indicating that the extensive use of natural wood in the elevated street and alley design substantially enhanced users’ emotional experience through its capacity to evoke natural affinity and perceptual warmth. This finding aligns with urban vitality studies emphasizing how material coherence in street-scale environments fosters emotional engagement, yet the consistent strength across multiple material dimensions exceeds conventional focus on single aesthetic elements in streetscape design. The minimal role of spatial privacy contrasts with commercial complex studies that often prioritize secluded areas, instead supporting street vitality research where visual accessibility enhances perceived safety and social interaction.

THE BOX (rank 2) shows minimal score difference from the top-rated case (0.659 vs. 0.660) but demonstrates slightly weaker emotion-promoting effects overall (Figure 4). The homogeneous contributions from its material and non-material features reflect the complex’s high functional richness and spatial flexibility, which provide diverse user groups with multiple emotional engagement pathways. This pattern resonates with mixed-use complex studies where program diversity distributes emotional impacts across various features, yet challenges street-scale research that typically attributes emotional responses to distinct physical elements. The non-significant relationship between color and positive emotions contradicts conventional urban design theories linking chromatic variety to attraction, instead aligning with emerging evidence that visual complexity in confined commercial spaces may induce perceptual overload.

Beijing Fun (rank 3) maintains strong attachment levels (score = 0.636) while uniquely suppressing negative emotions through both material features and non-material dimensions of sociability and uniqueness (Figure 5). The significant intercorrelations among color, form, and structure, along with their connections to regional identity, validate the effectiveness of preserving historical architectural elements in creating emotionally engaging environments. This multi-dimensional negative emotion suppression extends beyond typical commercial complex studies that focus primarily on positive emotion promotion, while the material-cultural coherence finding deepens street-scale heritage research by quantifying how traditional elements contribute to contemporary emotional experiences.

751D·PARK (rank 4) achieved moderate-to-high evaluation scores with non-material features exerting stronger influence than material characteristics (Figure 6). The unique positive correlation between playfulness and negative emotions represents a distinctive case where user expectations about interactive engagement with industrial heritage elements were not met through the current visual-focused design approach. This finding challenges conventional vitality studies in both street and complex contexts, where playfulness typically correlates with positive affect, instead revealing how symbolic preservation without functional adaptation can create emotional dissonance—a phenomenon observed but rarely quantified in previous industrial heritage research.

Galleria-Instreet (rank 5) received significantly lower evaluation scores despite both material and non-material features contributing to attachment formation (Figure 7). The significant suppression of negative emotions through non-material features did not translate to strong overall attachment, while the visually dominant design approach limited opportunities for sustained emotional engagement. This dissociation between negative emotion reduction and attachment formation challenges environment-behavior theories in both street and complex studies, where stress reduction typically correlates with place bonding. The design’s emphasis on visual enhancement over experiential engagement reflects a common limitation in contemporary commercial complex design, while the impact of non-operational commerce reinforces economic vitality as fundamental to emotional attachment in urban spaces.

Based on the foregoing analysis, the key factors and underlying mechanisms responsible for the attachment levels across the five vibrant complexes have been preliminarily clarified, which also validates the effectiveness of the evaluation framework developed in this study. That said, spatial vitality is inherently dynamic and evolving, susceptible to influences from temporal trends, social dynamics, and surrounding contextual developments. Therefore, the assessment of vibrant spaces from an emotional attachment perspective requires comprehensive interpretation, wherein qualitative approaches such as on-site observation and interviews remain an indispensable supplement. Meanwhile, the comparative analysis with previous studies reveals that successful emotional engagement in urban vitality spaces requires balancing material authenticity with functional adaptability across different scales. While street-scale environments benefit from material coherence and visual accessibility, commercial complexes require deeper integration between physical features and social programming to sustain emotional connections.

6. Conclusions

In summary, the transformation of buildings and urban spaces requiring renovation into commercially oriented mixed-use complexes, coupled with fostering users’ emotional attachment to them, represents an effective human-centered approach to enhancing vitality within the context of urban stock renewal. Addressing the current research gaps in quantitative studies on small-scale vibrant commercial spaces and the general neglect of emotional dimensions in urban vitality research and practice, this study develops and empirically validates a methodological framework applicable to the study and evaluation of commercial urban vibrant spaces through five recently completed representative complexes in Beijing. Based on the systematic construction, application, and analysis of this evaluation framework, the research contributes to urban design and research practice through three interconnected dimensions:

Theoretically, it advances urban vitality researches by integrating emotional attachment theory with space design, providing new insights into how people-environment relationships evolve in renewed urban contexts. According to the evaluation results of five commercial complexes, several insights emerge regarding the formation of attachment between people and vibrant urban spaces. Material features, such as colors, materials, and form and structures, promote emotional attachment significantly. Preserving authentic materials and incorporating those with historical meaning help people connect with a space through multidimensional experience including memories. This pattern holds across both street-level spaces and larger complexes, though each context requires different design strategies. While natural elements are recognized for supporting attachment, they often serve decorative rather than experiential roles. In terms of non-material features, including sociability, playability, and uniqueness, their influence appears more variable. Privacy, for instance, demonstrated limited impact across the five complexes. Non-material traits played a secondary role compared to physical ones. In lower-ranking cases, these features helped mitigate negative emotions but did not enhance overall attachment levels, suggesting they cannot independently sustain vibrant spaces. Overall speaking, material authenticity serves as the primary driver of emotional attachment. Non-material attributes provide complementary value, yet their effectiveness depends on the presence of a meaningful built environment.

Methodologically, the results introduce a quantitative framework that combines attachment scales with multi-criteria decision analysis to evaluate vibrant commercial environments. The integrated approach offers practical tools with clear application potential. The emotional attachment scale method provides a structured way to study small-scale spaces where traditional big data approaches fall short. This helps designers incorporate emotional evidence alongside functional and aesthetic considerations. The enhanced TOPSIS model, integrated with entropy weighting and grey relational analysis, creates a more objective evaluation system. It calculates weights without subjective bias, addresses information gaps in traditional models, and generates clear rankings for decision-making. While developed for commercial settings, the framework has the potential to adapt well to other contexts. It scales from individual buildings to district-level analysis, applies to cultural sites and transport hubs, supports both one-time and longitudinal assessment. With cultural calibration, it could work across international contexts. By linking emotional experience with quantitative analysis, the method creates a reproducible protocol that helps researchers and designers develop more engaging spaces in urban renewal contexts.

Practically, the research provides practical guidance for urban planning and design that can be applied across different regions and scales. For Chinese cities, it offers a framework to assess emotional attachment to develop locally tailored designs while integrating successful elements from other contexts. Planners can use this as a diagnostic tool to evaluate existing areas and set up monitoring systems for urban renewal projects. Meanwhile, the approach supports evidence-based design across various urban scales. For example, at the city level, emotional attachment metrics can be incorporated into broader vitality evaluation. And for smaller districts, it helps coordinate material and social features across connected spaces. Street-level applications give specific guidance on combining colors, materials, and natural elements with social qualities to boost engagement. For designers working on historical areas, the study suggests using materials that authentically convey historical stories while creating social connections that support contemporary use. In new developments, combining innovative new materials with adaptable social features helps establish unique identity. Besides, post-occupancy evaluations using this method allow continuous refinement of designs based on how people actually form emotional connections. This gives urban professionals practical tools that link spatial designs to emotional outcomes.

Though this study moves forward how we assess emotional attachment in urban commercial spaces, it’s important to recognize its limitations to point to its future prospects. On the one hand, the complex math methods represent both an advance and a challenge. The technical sophistication could overshadow the actual human experiences the study trying to understand. Though the data helps compare spaces systematically, but they can’t fully capture the qualitative ways people bond with places. Future work would mix these quantitative measures with deeper qualitative approaches to emphasize human experience. On the other hand, the findings are context-specific to some extent. The emotional attachment patterns identified in Beijing’s commercial spaces emerge from particular Chinese urban conditions and cultural dynamics. How these translate to western cities or different urban contexts remains an open question. Future research should test these indicators across cultures, identifying what’s universal versus what’s culturally particular when it comes to emotional attachment generated from urban spaces. Meanwhile, these limitations open useful paths forward. The methodological questions call for approaches that balance data with human experience, while the contextual constraints highlight the need for culturally sensitive evaluation tools such as ergonomic techniques. Addressing these challenges can help develop more nuanced ways to understand emotional vitality across increasingly diverse urban world.