Evaluating the Whole-Process Management of Future Communities Based on Integrated Fuzzy Decision Methods

Abstract

Focusing on the potential risks involved in the entire process of developing and creating Future Communities for old and renovated developments, a Future Community evaluation system for life-cycle management was constructed, featuring ten aspects. The index is preliminarily screened through the Fuzzy Delphi Method (FDM). The index weight is produced using the comprehensive weighting model of game theory based on the G1 method and Entropy Weight method. Six sample communities are ranked based on the TOPSIS method. This article proposes a feasible evaluation system for the comprehensive process of developing Future Communities. The results can provide relatively scientific evaluation results regarding the entire process of Future Community development, while promoting their sustainable operation and providing reference for other community regeneration projects in Zhejiang Province, China, and worldwide.

1. Introduction

As China enters a new era of urbanization, the transformation of its aging neighborhoods presents a multitude of uncertainties and complexities. The development of Future Communities represents a “new path” being actively explored in Zhejiang Province in China to overcome the challenges faced during this stage of urbanization. To date, seven rounds of successful declarations have resulted in 1263 Future Community projects. However, issues have arisen throughout the creation and operation of these Future Communities due to policy changes, conflicting interests, and insufficient planning. This is particularly evident in Future Communities undergoing numerous kinds of regeneration which demand a focused effort on fine-tuning and comprehensive management throughout the entire process.

Future Communities must confront the challenge of achieving long-term sustainable operation amidst the rapid proliferation of numerous Future Communities. Their swift expansion exacerbates issues related to limited spatial resources and the residents’ limited financial capabilities. Consequently, the commercial and service sectors within these communities tend to lean towards public welfare and micro-profits, necessitating substantial government subsidies for the construction of facilities and comprehensive service coverage. This dependence on government support complicates the accomplishment of balanced operating funds and the provision of sustainable services in the later stages of development. Consequently, Future Communities in regeneration areas are under significant pressure to maintain long-term viability. The multifaceted issues of financial equilibrium and sustained service provision underscore the formidable challenges inherent in the ongoing operation of Future Communities.

Indeed, the entire process of creating Future Communities reveals several specific challenges in terms of operation and management coordination. For instance, during the initial planning phase, policy thresholds often hinder community declarations. Information asymmetry leads to discrepancies between residents’ wishes and community planning, with insufficient consideration given to the diverse perspectives involved. The concept of “front-end operations” remains unexplored, and spatial planning tends to be formalized without due consideration of practical functionalities. The scale of these communities may at times be excessively large or dispersed, hindering the efficient deployment and accessibility of public service facilities. Additionally, research into demand may be inadequate. In the mid-term and construction stages, communities frequently grapple with funding shortages, resulting in deviations from intended spatial planning outcomes. In the later operational stage, issues such as a lack of integration between the Wisdom platform and other provincial platforms may lead to redundant constructions. Digital platforms may go unused, and the actual utilization rate of public facilities may be suboptimal. Moreover, insufficient operating funds pose challenges to the sustained operation of communities. In summary, imperfections in the processes of Future Communities’ creation, planning, and evaluation, coupled with residents’ reluctance to participate, the absence of front-end operations, and the disparity between operational situations and creation goals, underscore the need to enhance existing decision-making and evaluation processes to better align with the practical demands of the work at hand.

Furthermore, there is a need to regulate the value orientation of community renewal and construction. The primary aim of Future Community development is to de-emphasize real estate, all while enhancing residents’ well-being, without increasing their financial burden. However, the prevailing reality indicates that some Future Community land transfer prices exceed the norm. Consequently, the long-term supervision and management of Future Communities have become critical tasks. Our systems of evaluation for Future Communities, serving as foundational and managerial tools for operations, require in-depth study and refinement.

Finally, Future Communities stand at a pivotal juncture in their development, poised for full-scale promotion across the province and the nation post-2023. To ensure responsible and sustainable growth, it is imperative to institute corresponding institutional mechanisms that will regulate their expansion. Establishing a rationalized and standardized assessment mechanism and drawing from the lessons of existing pilot programs become crucial. Implementing a refined management approach, characterized by “relaxing declarations and increasing assessments”, will be instrumental in overseeing the declaration, construction, and operation of transformed Future Communities. This strategic move aims to strike a balance between encouraging development and maintaining stringent assessments of sustained quality and effectiveness.

In this area, the Building Research Establishment Environmental Assessment Method (BREEM) evaluation system stands as one of the earliest and most influential models for community evaluations, encompassing social, environmental, and economic sustainability goals, as well as policies. Over time, region-specific evaluation methods have emerged that are tailored to the unique circumstances of each locality, such as HK-BEAM in Hong Kong and BREEAM in Canada [1]. The U.S. Green Community Certification (LEED-ND) distills years of theoretical research into actionable guidelines, striking a harmonious balance between scientific rigor, practicality, and usability [2]. In China, the development of community evaluation systems has a rich history. Its origins can be traced back to the work of Taiwanese scholar Hsiao Tsai-an in 1992, who employed expert group decision-making models and Analytic Hierarchy Process (AHP) methodologies to reconcile and synthesize divergent expert opinions. This approach aimed to facilitate the optimization and enhancement of decision making for the selection of public facility locations [3]. In the same year, Zeng introduced brainstorming, a group decision-making technique, to construct a system for evaluating the environmental quality of metropolitan areas [4]. Tu (2014) pioneered the application of “group decision making” in architectural planning, particularly for large and complex projects. He established a “three-level feedback” model, integrating data from the decision-making body and object; and leveraged decision-making plug-ins in Building Information Modeling (BIM), SPSS 26 software, and other computer technologies to collect group preferences and analyze data outputs. These scores served as the foundation for subsequent projects’ implementation and the evaluation of the current status of existing projects [5]. Evaluation systems for other communities around the globe often center on various aspects, such as ecological health [6], community resilience [7], community vitality [8], neighborhood vitality [9], community public service facilities, child-friendly communities, and age-appropriate retrofitting [10]. In the case of Future Communities, the official evaluation system primarily emphasizes community resilience and neighborhood vitality. Zhejiang Province’s official evaluation system for Future Communities is articulated in the “Acceptance Measures for Future Communities in Towns and Cities in Zhejiang Province”. This comprehensive framework is used to evaluate the success of Future Communities through establishing indicators for nine different scenarios alongside characteristic highlights that shape a given Future Community. Importantly, it covers most of the areas that have been extensively researched and provides a comprehensive framework for assessing Future Communities [10].

However, the evaluation of Future Communities is not without limitations. According to the Management Measures for the Pilot Construction of Future Communities in Zhejiang Province, competent authorities currently focus their evaluation and assessment on the stages of program declaration and final acceptance. The Evaluation Indicator System for the Creation of Future Communities in Zhejiang Province, issued in 2021, serves as the basis for this evaluation, employing nine major scenarios and systematically assessing the outcomes of Future Community creation. The comprehensive evaluation system comprises nine major areas and 33 secondary indicators, offering control and guidance for the construction of Future Communities through binding and guiding indicators. During the acceptance stage, construction and operational effects are quantitatively evaluated, using the “Acceptance Measures for Future Communities in Towns and Cities of Zhejiang Province (for Trial Implementation)”, utilizing a scoring system of 520 points with an additional 120 points. To gain acceptance, a new community must achieve 500 points, while an old community must attain 360 points. However, the complexity induced by the more than 66 dimensions considered in the process significantly hampers the operationalization of the evaluation criteria. While these acceptance measures address operational acceptance, Future Communities that successfully gain acceptance and are listed still encounter challenges regarding their long-term sustainability. Therefore, it is imperative to adopt a holistic approach to Future Community management, incorporating front-end operation throughout the entire process. This inclusion is beneficial for achieving genuine, long-term sustainability and community development.

Consequently, this paper utilizes the pertinent theoretical framework of the life-cycle assessment (LCA) as its foundation. By simplifying post-assessment indicators and incorporating elements such as residents’ satisfaction and front-loaded operation, the aim is to construct a comprehensive evaluation system for Future Communities. This approach involves a meticulous review and organization of actual work processes to ensure that the evaluation system aligns more closely with the practical aspects of Future Community development.

In this paper, a viable evaluation system for the entire development process of Future Communities is proposed, encompassing ten dimensions: policy transmission, potential evaluation, residents’ demand, front-end operation, spatial scenarios, continuous operation, multiple subjects, financial balance, model optimization, and intelligent operation. The outcomes of this proposed system may be used to furnish relatively scientific evaluations of the comprehensive process of Future Communities’ creation and operation. Simultaneously, it is anticipated that the findings will contribute to the promotion of sustainable development in Future Community operations, offering valuable insights for the development of Future Communities, not only in Zhejiang Province but also in China and globally. Section 3 of this paper outlines the research design, introducing a comprehensive fuzzy evaluation method constructed through the integration of the Fuzzy Delphi Method (FDM), the Sequential Relationship Analysis Method (G1 Method), the Entropy Weighting Method, and the Approximation of Ideal Solutions Method (TOPSIS). Moving on to Section 4, a method for the evaluation and management of the whole process is applied. Drawing from research, interviews, and the literature, the entire process and potential risks associated with creating Future Communities are identified. Furthermore, a Future Community evaluation system, oriented towards holistic process management, is formulated. The indicators within this system are optimized using the Fuzzy Delphi Method (FDM). In Section 5, a game theory comprehensive assignment model based on a sequential relationship analysis (G1 method) and the entropy weight method is employed to assign weights to the indicators. The six sample communities are then ranked using the Approximation of Ideal Solutions Method (TOPSIS). The results are analyzed to highlight the strengths and weaknesses of the sample communities throughout the entire construction process.

2. Literature Review: Whole-Process Evaluation in Urban Planning Based on Fuzzy Multi-Criteria Decision-Making Methods

The primary objective of process management in Future Communities is to enhance quality of life and adherence to environmental regulations. A wealth of research addresses re-engineering issues in smart cities and Future Communities, employing innovative concepts, technologies, and methods. For instance, LM Olıverı et al. (2023) proposed an innovative method utilizing Industry 4.0 monitoring and control technologies to enhance the sustainability of composting processes, thereby creating new opportunities for Smart Cities [11]. Additionally, Fedorczak-Cisak et al. (2023) introduced a Smart City solution based on the establishment of self-sustaining housing communities with active user participation [12].

The concept of the “whole process” first emerged in the United States during the 1960s and 1970s and was initially referred to as the “Cradle to Grave” concept. In 1990, the Society of Environmental Toxicology and Chemistry (SETAC) introduced the concept of the life-cycle assessment (LCA), which consists of four interconnected components: defining objectives and determining scope, an inventory analysis, an impact assessment, and an improvement assessment [1]. Liberacki et al. (2023) conducted an examination of the Environmental Life-Cycle Costs (ELCCs) of Urban Air Mobility (UAM) as a crucial input for sustainable urban mobility [13].

Over the years, the concept of the whole process has evolved, and its application has been extended to the field of planning. In planning practices, the whole-process assessment method is commonly used in activities such as brownfield regeneration and the renewal of old cities. For example, Slagastad (2012) employed scenario construction and whole life-cycle assessment to evaluate the environmental benefits of different infrastructure construction options in the planning of new urban settlements [14]. Susca (2020) applied LCA concepts and methods to study the heat island effect in cities [15]. Salvati (2022) and others examined anomalies in the metropolitan area of Athens, Greece, over the entire life cycle and sought to explain local population growth patterns within this framework [16].

Currently, there is growing emphasis on full-cycle management in planning in China, from national land spatial planning to detail planning and special planning. This includes establishing a system to trace the entire process of planning preparation, approval, modification, and the supervision of implementation, which has since become part of inspections of natural resources law enforcement. In academia, community evaluations and planning evaluations based on whole-process management or life-cycle management have gained traction. For instance, Zhao et al. (2010) explored the application of life-cycle theory in urban and rural planning, outlining an urban-planning life-cycle system and identifying the characteristics of different stages of the urban life cycle and planning life cycle [17].

In recent years, the whole-process approach has permeated various planning subfields. Zou et al. (2011) conducted a whole-process study on mass participation in Shenzhen’s urban master plan, demonstrating a comprehensive approach involving public opinion surveys and the handling of public opinions [18]. Liu et al. (2021) established a whole-process working mechanism for community sports parks in Guangdong Province, addressing six key aspects: the planning, siting, design, construction, management, and operation of such facilities [19]. Chen et al. (2013) proposed a whole-process system framework encompassing four phases (engineering decision making, design, construction, and operation and maintenance), using a step-by-step independent evaluation method to ensure each phase’s completion and qualification before moving on to the next [20]. Xue et al. (2016) discussed the whole-process management of urban design, drawing from their practice and work experience in the planning industry [21]. Xiong et al. (2022) devised a multi-level renewal strategy, a multi-departmental linkage guarantee mechanism, and a departmental performance and social performance evaluation mechanism through a full-cycle perspective when renewing old urban areas. However, their specific evaluation dimensions are limited to planning and design, the guarantee of implementation, and the evaluation of performance, neglecting the evaluation of planning, operation, and construction [22].

Collectively, these studies have systematically disassembled and organized the internal operational mechanisms of communities and programs at each stage, while also identifying evaluation criteria based on relevant national standards and practical objectives. Through the integration of multi-subject and multi-criteria decision-making methods, localized indicator systems have been established. These endeavors have laid a vital theoretical foundation for the whole-process management of community renewal and planning.

Building planning seeks to employ a range of mathematical and scientific tools to create decision models, decision systems, and indicators that can be used to address various challenges in the building construction cycle, while accommodating individual and group preferences [23]. The field of multi-criteria, multi-subject decision-making research largely encompasses projects addressing the issue of groups with differing opinions struggling to make rational decisions [24].

In the 1980s, Bellman et al. recognized the necessity of specialized quantitative methods to quantify decision making in complex or uncertain environments and introduced fuzzy mathematics into multi-criteria decision making [25]. This introduction of fuzzy mathematics into the subject of multi-criteria decision making marked a pivotal turning point. Subsequently, multi-criteria decision making has found applications on a significant scale in the construction and planning industries. Exploration of the pre-planning and post-assessment of the built environment constitutes the subfield of multi-criteria and multi-subject decision making within the planning domain.

Today, there are several commonly employed tools for constructing community evaluation systems, including TOPSIS, DEA, AHP, MCDM, SD, GD, ISM, FDM, and group decision-making methods [26]. While these were initially applied in other disciplines, such as nursing and sociology, the inherently fuzzy and multi-subjective nature of community-related work has made these methods widely applicable in the field of community planning as well. Elagouz et al. (2023) introduced a hybrid life-cycle sustainability assessment (LCSA) model that integrates a multi-region input–output analysis with innovative multi-criteria decision-making techniques. Their approach is based on the Interval-Valued Neutrosophic Fuzzy (IVNF)–Analytic Hierarchy Process, with a Combined Compromise Solution (CoCoSo) approach [27]. In a similar vein, Li et al. (2023) developed a Location Selection Model using a Hybrid Fuzzy Decision-Making Approach. This model incorporates interpretive structural modeling and the fuzzy decision-making trial and evaluation laboratory (ISM-FDEMATEL) [28].

3. Research Design and Methodology: Fuzzy Integrated Decision-Making Method

3.1. Methodology for Establishing the Theoretical Framework for Evaluation: Inventory Analysis Based on LCA and BEQUEST

The methodology for establishing the theoretical framework for evaluation begins with a foundation in the life-cycle assessment (LCA) principles, which define the scope of the entire process of developing a Future Community. The entire process of the Future Community’s development is then systematically organized through a checklist analysis, creating a theoretical framework for the evaluation index system. Various techniques, such as checklists, flowcharts, brainstorming, and scenario analysis, are often used in combination for specific applications [29,30].

In this context, said checklist analysis is integrated with an analysis of the built environment. Furthermore, the checklist analysis is combined with the Built Environment Sustainability Evaluation Framework (BEQUEST) to identify the input and output elements of the system. Data collection takes place during the process of the indicators’ establishment and the evaluation’s application.

3.2. Optimizing Evaluation Indicators: The Fuzzy Delphi Method (FDM)

The Delphi Method, also known as the Expert Correspondence Method, is a problem-solving approach based on the way that individuals think and respond to uncertainty and ambiguity in reality, drawing from human experiences and expertise [31]. The Fuzzy Delphi Method is an adaptation of the Delphi Method that incorporates fuzzy theory. It enables experts to provide opinions with a higher degree of uncertainty, particularly when addressing problems that are challenging to quantify or inherently subjective in nature.

3.3. Determining Indicator Weights: Subjective and Objective Game Theory Assignment Combining the G1 and Entropy Weight Methods

The G1 method, known as the sequential relationship analysis, is an enhancement of the Analytic Hierarchy Process (AHP), which falls under the category of subjective weighting methods. AHP, a versatile data analysis tool that effectively blends qualitative and quantitative factors, has found widespread use across various domains. However, AHP faces limitations in practice, such as difficulties in achieving consistency within the judgment matrix, complex processes, and challenges when dealing with numerous indicators [32]. In contrast, the G1 method, an improvement on AHP, offers greater ease of operation and eliminates the need to address the consistency factor [33,34].

Entropy weight, originating from the field of physics, is based on the concepts of information and entropy. Information measures the degree of order within a system, while entropy quantifies the degree of disorder. The combination of these concepts forms the basis of information entropy. According to the definition of information entropy, an indicator’s entropy value reflects the degree of its dispersion. Smaller entropy values suggest greater impact and influence on comprehensive evaluation (i.e., weight). Thus, information entropy serves as a tool for calculating the weight of each indicator, providing a foundation for comprehensively evaluating multiple indicators. The entropy weighting method offers several advantages, including dimensionless appropriateness, robustness, monotonicity, and independence from scaling. Additionally, it endeavors to derive optimal weights based on objective and real data, comprehensively and realistically reflecting the information contained within the indicator data. However, this method is highly dependent on sample data, which can introduce biased results in certain situations. For instance, if all values of an indicator are equal, that indicator will not influence the comprehensive evaluation.

Game Theory Combinatorial Empowerment, a fusion of game theory and combinatorial optimization, primarily focuses on decision-making problems in multi-player games. It abstracts the participants, decisions, and benefits in a multi-player game into a mathematical model and employs combinatorial optimization techniques to solve it. In this context, participants can be individuals or organizations, each assigned a decision weight and benefit weight. Decision weight signifies decision-making ability, and benefit weight represents the degree of benefit. Participants can choose different strategies during decision making, each leading to varying outcomes, which can be positive, negative, or neutral. By assigning weights to all possible strategies and combining them, an optimal decision-making scheme is determined to maximize benefits for all participants.

3.4. Information Clustering Evaluation Methodology: The Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS)

The Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) is a comprehensive evaluation method that is widely employed within group settings. It fully utilizes original data and provides results that accurately reflect the differences between evaluation programs. The fundamental process involves normalizing the original data matrix and employing the cosine method to identify the best and worst solutions among a limited set of options. Subsequently, the method calculates the distance between each evaluation object and the best and worst solutions, respectively, to determine the relative proximity of each evaluation object to the best solution. This relative proximity serves as the basis for evaluating advantages and disadvantages. Our research design, based on this method, is illustrated in Figure 1.

4. Future Community Evaluation for a Whole-Process Management Indicator System

Upon scrutinizing the entire life cycle of Future Communities, this study discerns that the pre-planning phase, as the foundation, significantly shapes the trajectory and state of the community’s creation and operation. The operational challenges encountered later often stem from issues that are firmly established during the creation phase. Therefore, increasing emphasis on operational considerations and residents’ preferences during the pre-planning stage can effectively mitigate future operational difficulties. In light of this, this paper champions a holistic approach to whole-process management and front–back synergy, offering a comprehensive evaluation system with a framework that includes the following layers: planning linkage, current status assessment, residents’ preferences, front-loaded operation, spatial intensity, scenario placement, ongoing operation, multiple stakeholders, financial equilibrium, model refinement, and intelligent operation (as depicted in Figure 2). Notably, this approach omits the creation phase as a separate evaluation dimension, opting to view the entire process as an integrated whole, thereby eliminating redundancy in the evaluation system between the initial and middle phases of the process.

4.1. Establishment of a Preliminary Indicator System

Our selection of indicators adheres to the principles of scientific validity, comprehensiveness, and measurability. Drawing on guidance from various sources, including the “Guidelines for the Preparation of Special Planning for Urban Community Construction in Zhejiang Province (for trial implementation)”, “Technical Points for the Declaration and Design of Future Communities”, “Measures for the Acceptance of Future Communities in Towns and Cities in Zhejiang Province”, “Measures for the Management of Pilot Construction of Future Communities in Zhejiang Province”, relevant meetings, planning documents, and an extensive review of planning reports and research studies conducted in the 44 districts and communities of Zhejiang Province, a preliminary evaluation system was developed through a combination of an analysis of the literature and research interviews. This preliminary system is presented in

Table 1. System of indicators for the whole-process evaluation of Future Communities.

Normative Hierarchy	Indicator Layer/Reference Source	Normative Hierarchy	Indicator Layer/Reference Source
Policy transmission	Participation in sectoral linkages	Scene location	Scene selection rationality
	Community boundary harmonization		Scene landing completion
	Degree of implementation of planning transmission		Characteristic highlights shaping the degree
	Completeness of the supporting construction guidelines	Ongoing operation	Operational service perceptibility
Evaluation of potential	Community road network density suitability		The scene continues to open up
	Degree of sophistication of community public service facilities	Pluralistic subjectivity	Long-term stability of the operating entity
	Community size suitability		Degree of diversification of governance actors
	Adequacy of community population size		Resident satisfaction with multi-principal governance
Residents’ willingness	Resident needs survey	Cash balance	Working capital closure
	Residents’ willingness to create		Marketization of public service facilities
	Willingness to build communities		Diversification of funds
Front-end operations	Functional space planning and design front	Mode optimization	Transportation and management merchants’ quality degree
	Operator planning and design engagement		Optimization of community operation dynamics
Spatial integration	Compound use of space		Degree of digital platform intensification
	Stock space utilization	Intelligent operation	Effective operation of digital platforms
	Sharing of public service facilities		Digital operational satisfaction

.

4.2. Indicator Screening Based on the Fuzzy Delphi Method (FDM)

To refine the indicators within our whole-process evaluation system for Future Communities, this paper employs a modified Fuzzy Delphi Method. The process begins with the creation of the first round of expert questionnaires, which are based on the initial Future Community whole-process evaluation index system proposed in Section 4.1. These questionnaires are further informed by insights gathered from the literature research, industry interviews, and on-site investigations.

The initial index system is iteratively refined through expert consultations over two or three rounds, modifying the system until the opinions of experts converge. A threshold value is established to remove indicators that fail to converge, ultimately producing a formal version of our whole-process evaluation and management indicator system for Future Communities.

To ensure the scientific validity of the research, 12 experts with direct experience in Future Community-related work were invited to form an expert panel. Unlike within the traditional Delphi method, which requires a recalculation of the positive coefficients of experts, the prior formation of the expert panel obviates this step. After each round of the process, the scientific validity of the results is confirmed via the calculation of the average scores assigned to each indicator by the experts in the decision-making group, the rate of unanimous agreement, and the level of consensus among the experts.

The indicators are screened based on the average scores from invited experts and the coefficient of variation. Indicators are categorized according to their importance, with ratings ranging from 5 (very important) to 1 (unimportant). Indicators with average scores below 3 are eliminated, while those with scores greater than 3 and coefficients of variation less than 0.2 are retained.

If we recognize that the preliminary indicator screening may not cover all possibilities, adjustments can be made based on expert input, including the addition or removal of indicators, provided that the experts offer reasoned justifications and a consensus among most experts is reached. The survey is conducted in two stages.

4.2.1. First Round of Indicator Screening

Following the completion of expert scoring, the questionnaires are collected, and a data analysis is conducted to calculate the scores assigned to the indicators. During this process, indicators that fail to meet the specified criteria are removed. Additionally, the indicator system is adjusted in response to suggestions made by the experts regarding the addition or removal of indicators. This may include the potential inclusion of new indicators or modifications to the significance levels of certain indicators.

In the first stage of the survey, ten experts from the relevant fields within the expert panel were contacted via correspondence. A total of 12 questionnaires were distributed, with all 10 returned questionnaires deemed valid. The response rate of the experts was 83%.

In the initial analysis, the experts’ familiarity with the content of the indicator system (Cs) and their coefficient of judgment (Ca) were assessed. The assigned values for each aspect are provided in

Table 2. Expert familiarity (A) assignment.

Familiar	More Familiar	General Familiarity	Less Familiar	Unfamiliar
1	0.8	0.5	0.2	0

and

Table 3. Expert judgment factor.

Basis of Judgment	Degree of Influence on Expert Judgment
	Oldest	Center	Few
B1, theoretical knowledge	0.3	0.2	0.1
B2, practical experience	0.5	0.4	0.3
B3, peer understanding	0.1	0.1	0.1
B4, subjective perception	0.1	0.1	0.1

[28].

According to the results of the first round of correspondence, the experts’ familiarity and judgment coefficients are shown in

Table 4. Expert panel familiarity and judgment factor.

	Expert 1	Expert 2	Expert 3	Expert 4	Expert 5	Expert 6	Expert 7	Specialist 8	Expert 9	Specialist 10
Familiarity	Familiar	More familiar	More familiar	More familiar	Familiar	More familiar	Familiar	More familiar	Familiar	More familiar
Theoretical knowledge	Center	Center	Your (honorific)	Your (honorific)	Your (honorific)	Your (honorific)	Your (honorific)	Center	Your (honorific)	Center
Practical experience	Your (honorific)	Center	Your (honorific)	Center	Center	Your (honorific)	Your (honorific)	Center	Your (honorific)	Your (honorific)
Peer understanding	Your (honorific)	Center	Your (honorific)	Center	Your (honorific)	Your (honorific)	Your (honorific)	Your (honorific)	Center	Center
Subjective perception	Center	Center	Your (honorific)	Center	Lower (one’s head)	Your (honorific)	Your (honorific)	Center	Lower (one’s head)	Center

. It is notable that all experts in this round achieved a high degree of expert authority (Cr), with a value of 0.88. Typically, when the Cr is equal to or greater than 0.7, it signifies a favorable correspondence outcome, indicating that the data from this expert correspondence questionnaire are highly credible.

To gauge the degree of experts’ consensus, the coefficient of variation (Cv) is commonly used. The Cv serves as an indicator of the level of consensus among experts regarding the importance of each indicator, the rationality of the calculation formula, and the practicality of the data collection method. A smaller Cv value indicates a higher degree of consensus among experts. The expert evaluations and the Cv values for each indicator in this round are presented in

Table 5. Summary of data processing for indicator screening.

Standardized Layer	Indicator Layer	Results of the First Round of Data Calculations for Indicators			Results of the Second Round of Data Calculations for Indicators
		Mean, Mj	Standard Deviation, Aj	Coefficient of Variation, Cv	Mean, Mj	Standard Deviation, Aj	Coefficient of Variation, Cv
Policy transmission	Participation in sectoral linkages	4.3	0.640	0.149	4.7	0.458	0.098
	Community boundary harmonization	3.4	0.917	0.270	3.9	0.831	0.213
	Degree of implementation of planning transmission	3.5	1.025	0.293	4.7	0.640	0.136
	Completeness of the supporting construction guidelines	3.2	1.327	0.415	3.6	0.663	0.184
Evaluation of potential	Community road network density	2.9	0.943	0.325	4	0.447	0.112
	Degree of sophistication of community public service facilities	4.5	0.500	0.111	4.8	0.400	0.083
	Community size suitability	3.8	0.748	0.197	4.6	0.490	0.106
	Adequacy of community population size	3.8	0.748	0.197	4.2	0.400	0.095
Residents’ willingness	Resident needs research Coverage	4.4	0.663	0.151	3.7	0.458	0.124
	Residents’ willingness to create	3.5	1.025	0.293	4.1	0.831	0.203
	Willingness to build communities	3.9	0.831	0.213	4.3	0.640	0.149
Front-end operations	Functional space planning and design front	4.5	0.671	0.149	4.3	0.781	0.182
	Operator planning and design engagement	4.6	0.490	0.106	4.2	0.748	0.178
Spatial integration	Degree of composite space utilization	2.9	0.831	0.286	3.6	0.663	0.184
	Stock space utilization	3.9	0.943	0.242	4.5	0.671	0.149
	Degree of sharing of public service facilities	4.1	0.831	0.203	4.6	0.490	0.106
Scene location	Scene selection rationality	3	1.414	0.471	4.2	0.980	0.233
	Scene landing completion	4.5	0.500	0.111	4.3	0.458	0.107
	Characteristic highlights shaping degree	3	1.414	0.471	3.9	0.700	0.179
Ongoing operation	Operational service perceptibility	4.2	0.400	0.095	4.2	0.600	0.143
	The scene continues to open up	4	0.632	0.158	4.0	0.632	0.158
Pluralistic subjectivity	Long-term stability of the operating entity	4.2	0.748	0.178	4.5	0.500	0.111
	Degree of diversification of governance actors	3.9	0.700	0.179
	Residents’ satisfaction with multi-principal governance	4.1	0.831	0.203
Cash balance	Working capital closure	4.2	0.872	0.208
	Marketization of public service facilities	4.2	0.600	0.143
	Diversification of funds	3.7	0.748	0.1972
Mode optimization	Transportation and management merchants’ quality degree	3.8	0.600	0.158
	Optimization of community operation dynamics	4	0.775	0.194
Intelligent operation	Degree of digital platform intensification	4	0.632	0.158
	Effective operation of digital platforms	4.3	0.640	0.149
	Digital operational satisfaction	3.6	1.020	0.283

.

Cv = A_j/M_j

(1)

Following the data processing of the statistical indicator system, the mean values of the indicators were computed. In combination with the criterion of Cv ≤ 0.2 for the average score, some indicators were considered for removal [35]. Consequently, the following indicators were excluded from the indicator system: the degree of coordination of the community boundaries, degree of implementation of the planning conduction, density of the community road network, willingness of residents to create, willingness to build community, composite use of space, use of space in the stock, reasonableness of scenarios, degree of characteristics of the highlights of shaping, digitalization, and satisfaction with digital operation.

Additionally, the expert panel suggested that the emphasis should not solely be focused on the implementation of Future Community planning and the self-governance of residents; rather, the key focus should revolve around operational self-consistency, encompassing aspects such as economic accounting, operational challenges, and the stability of Future Community operations.

4.2.2. Second Round of Indicator Screening

Following the data processing of the first round of indicators, the indicator system, post initial screening, was organized. Based on this refined system, a second round of expert questionnaires was prepared. The results of this second round were analyzed and processed accordingly.

In the second stage, revised questionnaires were distributed to the same 10 experts who had participated in the previous round. A total of 10 questionnaires were issued, with all 10 returned questionnaires deemed valid. The same analytical process was applied to the experts’ evaluation data, leading to the exclusion of indicators that did not meet the specified criteria. The results of the indicator calculations in both rounds can be found in Table 5. The statistical analysis of the data reveals that the mean value of all indicators exceeds 3, indicating that all indicators within the system are considered very important, and there is no need to proceed with another round of expert correspondence. However, upon observing the coefficient of variation for each indicator, it was noted that three indicators, specifically the improvement of supporting construction guidelines, the completion of scenario landing, and the marketization of public service facilities, have coefficients exceeding 0.2 and are thus temporarily excluded from the system. Additionally, following the Cronbach’s Alpha Reliability Test, the coefficient of 0.769 exceeded 0.7, signifying that the reliability of the collected data is good. Furthermore, the experts provided valuable suggestions in their questionnaires:

Experts in the field of engineering management emphasized that the construction of the system should encompass all aspects of the development of Future Communities, including planning, design, operation, management, capital operation, community governance, and service support. A comprehensive approach to these aspects may enhance the comprehensiveness, efficiency, convenience, and sustainability of Future Communities, allowing them to better meet the needs of residents in terms of livability, workability, and tourism.

Experts in architectural design have highlighted the importance of indicators related to the completeness of community public service facilities, resident demand research, front-end functional space planning and design, scenario completion, ongoing scene access, resident satisfaction with multi-entity governance, the dynamic optimization of community operations, and the effective functioning of digital platforms. They suggest that indicators such as the appropriateness of community area and the degree of diversification of governance subjects may carry relatively lower weights.

Urban planning experts stress the significance of aspects such as community size, the quality of public service facilities, functional space planning and design, and research into the demands of residents within community planning and construction. They also emphasize the need for both commercial and community operations operators in Future Communities. The goal of scenario construction is to facilitate the efficient sharing of scene resources among residents, thereby reducing the waste of public resources.

4.2.3. Optimization of the Evaluation Indicator System

Through interviews with experts, and after obtaining the unanimous agreement of all experts, the degree of operators’ participation in planning and design is modified to the operators’ degree of participation in pre-planning; the degree of the utilization of stock space is incorporated into the system again, and the degree of the completion of scenario landing is removed. The spatial intensification layer and scenario landing layer are merged into the spatial scenario layer; the degree of the guidelines’ suitability for the construction of ancillary facilities, the marketization of public service facilities, and the degree of completion of scenario landing are deleted to thusly form an optimized indicator system (

Table 6. Optimized system of indicators for the whole-process management of Future Communities.

Standardized Layer	Indicator Name	Connotation of Indicators
Policy transmission, A	Participation in sectoral linkages, A1	Number of participating departments and links in the whole process of creation
Potential evaluation, B	Community public service facilities improvement, B1	Ratio of public service facilities available before the creation of the community to those available after the creation of the community
	Community size suitability, B2	Implementation of unit size fit with future neighborhoods
	Community population size fitness, B3	Suitability of the size of the implementation unit to the population of the community
Demand of the population, C	Resident needs survey, C1	Coverage of research on residents’ needs
Front-end operations, D	Functional space planning and design front, D1	Number of facilities with operational functions pre-programmed in advance
	Operator pre-planning participation, D2	Whether the operator was involved during the planning period, and the number of sessions in which it was involved
Space scene, E	Degree of sharing of public service facilities, E1	Percentage of facilities within the community that are shared with the outside
	Stock space utilization, E2	Ratio of space area of the original facility to the area of the created facility
Continuing operation, F	Operational services perceptible, F1	Can the public perceive operational services?
	Scene continues to open, F2	Average daily open hours for all scenarios
Multi-subject, G	Long-term stability of operating entities, G1	Number of years the operator has been contracted
	Degree of diversification of governance actors, G2	Number of community-based grassroots governance groups
	Resident satisfaction with multi-objective Governance, G3	Residents’ satisfaction
Balance of funds, H	Working capital closure degree, H1	Are there any other fund balancing highlights for the community beyond the funding closure?
	Degree of diversification of funds, H2	Number of sources of funding for community functioning
Model optimization, I	Transportation merchants’ quality degree, I1	Number of quality stores (rated 4 stars) as a percentage of
	Community operations dynamics optimization, I2	The insufficiency of the operation process can be found in time, timely feedback and modification can be provided.
Intelligent operation, J	Degree of digital platform intensification, J1	Access to provincial digital platforms for Future Communities and integration of all functions.
	Effective operation of digital platforms, J2	Comprehensive use of community platforms in the last week

).

5. A Future Community Evaluation System for Whole-Process Management—Case Studies on the Ranking of Six Future Communities in Zhejiang Province, China

5.1. Determination of Subjective Weights Based on G1 Analysis

The difference between the G1 method and the AHP method is that G1 does not need a judgment matrix to be constructed; rather, it establishes the order of the relationship between the indicators to express the approximate degree of importance between them. Firstly, the order relationship is determined: Y₁ > Y₂ > Y₃ > …… Y_n. Then, the importance level between neighboring indicators is determined, assuming that the ratio of the importance level between neighboring indicators Y_k−1 and Y_k according to the subjective judgment of the experts is R_k = Y_k−1/Y_k, where k = 1, 2, 3 ……, m. Finally, the calculation of the subjective weight, a, is performed, and it should satisfy $a_k>0, {{\displaystyle \sum }}_{k=1}^m{a}_k=1$. The higher the value of R_K, the more important the indicator Yk−1 is compared to Yk.

$$a_k={\left(1+{{\displaystyle \sum }}_{k=2}^m{{\displaystyle \prod }}_{i=k}^m{R}_k\right)}^{-1}$$

(2)

$$a_{k-1}=R_K{a}_k$$

(3)

After the completion of the Delphi Expert Correspondence Questionnaire, two experts were re-invited to refer to the table and rank the importance of the indicators within the entire system. They were also asked to score the level of importance for each indicator.

Table 7. Sequential relationships among indicators.

Relationships among indicators	Y31 > Y21 > Y11 > Y41 > Y81 > Y73 > Y102 > Y61 > Y62 > Y82 > Y42 > Y71 > Y92 > Y101 > Y52 > Y23 > Y91 > Y51 > Y72 > Y22
Inter-indicator importance	1.0, 1.3, 1.0, 1.4, 1.1, 1.0, 1.2, 1.4, 1.4, 1.0,1.8, 1.4, 1.1, 1.2, 1.2, 1.2, 1.1, 1.0, 1.4, 1.2

shows the results that were obtained. The subjective weights of the indicators were calculated using MATLAB 26.

5.2. Determination of Objective Weights for Entropy Weights Based on Multi-Source Data

The determination of the objective weights for entropy weights based on multi-source data is a crucial step in the process of evaluating Future Communities. To ensure the reliability of the weights, it is usually recommended to use a sufficient number of samples (typically a minimum of five). In this project, six built and differentiated Future Communities were investigated, and the indicators were scored to calculate objective weights, using the entropy weight method. Additionally, the final weights were obtained based on game theory combination assignment and expert scoring assignment. This is a comprehensive approach that takes into account multiple factors and perspectives, therefore enhancing the robustness of the weight assignment process. The use of multiple samples, combined with entropy weight methodology and expert inputs, adds credibility to the weight assignment process, making it more suitable for evaluating the performance and characteristics of Future Communities. This comprehensive approach ensures that the weights are reliable and can be used for informed decision making in urban planning and development projects.

5.2.1. Community Selection

The identification of Future Communities involved a meticulous screening process through the government website directory and existing information sources. Six communities were selected based on the principles of data accessibility: the Jiangbin Future Community, Qibei Future Community, Zhongxing Future Community, Dohuli Future Community, Mutual Future Community, and Wangjiangyuan Future Community. Data for each indicator of the sample communities were comprehensively obtained through various channels, including each community’s governmental publication website; public resources trading center; community information; community planning documents; social media posts; and various types of mobile application data, map locations, and community research interviews. This comprehensive approach ensures a thorough understanding of the values associated with each indicator in the selected sample communities.

5.2.2. Scores and Objective Weights of Evaluation Indicators for Each Community

Most objective indicators were extracted from the target planning documents, while qualitative indicators, such as the perceptibility of operation services, residents’ satisfaction with the governance of multiple subjects, quality of transportation and management investment, dynamic optimization of community operations, and effective operation of the digital platform, pose a challenge for direct extraction. For the remaining objective indicators, big data crawling is employed. For instance, the indicator related to the quality of transportation and the promotion of investment is determined by the number of merchants in the area and the percentage of merchants with a rating of four stars or more on the Gaode map. The effective operation of digital platforms is assessed via the calculation of the total number of tweets from a given Future Community’s public account, considering both the total number of tweets and the highest number of views for tweets in a given week. Qualitative indicators, such as the perceptibility of operation services, residents’ satisfaction, and dynamic optimization, are derived from satisfaction surveys. For communities existing for more than a year, satisfaction survey results are directly obtained through community public channel tweets (e.g., Qibei and Wangjiangyuan). For newer communities without a satisfaction survey, a researcher-developed questionnaire is distributed through community properties and operators.

The initial step involves transposing the data and performing data quantification. Given the varied meanings and units of each indicator’s data, it is crucial to standardize the data initially to ensure the scientific integrity of subsequent calculations during statistical processing. For positive indicators, where a larger value signifies a better outcome, the positive nature of these indicators is maintained, and the data are normalized within a given range [0, 1]. Conversely, for reverse indicators, where a smaller value signifies a better outcome, inverse processing is applied, and the data are compressed within the [0, 1] range. Additionally, for indicators where there is an optimal point (meaning the closer the indicator is to a certain value, the better), data normalization within the [0, 1] range is applied. It is noteworthy that, with the exception of the degree of digital platform intensification, which is treated as a reverse indicator, the remaining indicators are considered positive indicators in this data collection (

Table 8. Dimensionality-normalized data.

Norm	Community 1	Community 2	Community 3	Community 4	Community 5	Community 6
a1	0.333	0.833	0	1	0.5	1
a2	0.365	0	0.794	0.205	0.837	1
a3	1	0.456	0.504	0.381	0.12	0
a4	0.492	0.25	0.103	1	0	0.352
a5	0.61	0.501	0.48	1	0.61	0
a6	0	1	0.444	0.722	0.5	0.944
a7	0	1	0.75	0.75	0.875	0.625
a8	0.19	0.335	0.659	0	0.485	1
a9	0.905	0.395	0	0.706	0.727	1
a10	0.429	1	0.857	0.143	0	0.286
a11	0	0.667	0.533	1	0.798	0.349
a12	0.333	0.667	1	0.667	0	0.667
a13	0	1	0.667	0.833	1	1
a14	1	0.333	0	0.183	0.444	0.808
a15	0	0.667	0.333	1	0.667	0.167
a16	0	0.333	0.333	0.667	1	0.333
a17	0.632	0.853	0.204	0	1	0.341
a18	0	0.875	0.625	1	0.25	0.75
a19	0	1	0.667	0.333	1	1
a20	0.769	0.521	0.179	0.752	1	0

).

$${\mathrm{b}}_{\mathrm{ij}}=\left(\mathrm{X-Min}\right)/\left(\mathrm{Max-Min}\right) \dots \dots \mathrm{forwarding}\left(i=1,2,3,\dots ;j=1,2,3,\dots \right)$$

(4)

$${\mathrm{b}}_{\mathrm{ij}}=\left(\mathrm{Max-X}\right)/\left(\mathrm{Max-Min}\right) \dots \dots \mathrm{Inverse}\left(i=1,2,3,\dots ;j=1,2,3,\dots \right)$$

(5)

After the data are made dimensionless, the data are shifted to ensure that they are all positive and do not affect the overall native state of the data.

$$x_{ij}=b_{ij}+0.001 \left(i=1,2,3\dots ;j=1,2,3,\dots \right)$$

(6)

The weight of the i^th sample indicator value in the j^th indicator is calculated as follow:

$$P_{ij}=\frac{x_{ij}}{{\Sigma }_i^m{x}_{ij}},0\le P_{ij}\le 1\left(i=1,2,3,\dots ;j=1,2,3,\dots \right)$$

(7)

A weighting matrix of the data is then obtained (

Table 9. Specific gravity matrix.

Pij	1	2	3	4	5	6
a1	0.091	0.227	0	0.273	0.136	0.273
a2	0.114	0	0.248	0.064	0.262	0.312
a3	0.406	0.185	0.205	0.155	0.049	0
a4	0.224	0.114	0.047	0.455	0	0.16
a5	0.191	0.156	0.15	0.312	0.191	0
a6	0	0.277	0.123	0.2	0.138	0.262
a7	0	0.25	0.187	0.187	0.219	0.156
a8	0.071	0.126	0.247	0	0.182	0.375
a9	0.242	0.106	0	0.189	0.195	0.268
a10	0.158	0.368	0.316	0.053	0	0.105
a11	0	0.199	0.159	0.299	0.238	0.104
a12	0.1	0.2	0.3	0.2	0	0.2
a13	0	0.222	0.148	0.185	0.222	0.222
a14	0.361	0.12	0	0.066	0.16	0.292
a15	0	0.235	0.118	0.353	0.235	0.059
a16	0	0.125	0.125	0.25	0.375	0.125
a17	0.209	0.282	0.067	0	0.33	0.113
a18	0	0.25	0.179	0.286	0.071	0.214
a19	0	0.25	0.167	0.083	0.25	0.25
a20	0.239	0.162	0.056	0.233	0.31	0

):

$$P=\left(\begin{array}{ccc}{P}_{11}& \cdots & P_{1n}\\ ⋮& \ddots & ⋮\\ P_{m1}& \cdots & P_{mn}\end{array}\right)\left(n=1,2,3\dots ;m=1,2,3,\dots \right)$$

(8)

The entropy value of the jth indicator, e_j, is defined as k = ln(t), where t is the number of samples, calculated entropy values are shown in

Table 10. Calculated entropy values.

Norm	a1	a2	a3	a4	a5	a6	a7	a8	a9	a10
ej	0.857	0.828	0.803	0.769	0.876	0.871	0.891	0.822	0.875	0.79
norm	a11	a12	a13	a14	a15	a16	a17	a18	a19	a20
ej	0.867	0.869	0.892	0.813	0.819	0.834	0.825	0.854	0.863	0.838

.

$$e_j=-k{{\displaystyle \sum }}_{i=1}^m{1}^1{P}_{ij}\ln P_{ij},k>0, 0\le e_{ij}\le z$$

(9)

Finally, the weights of the w_i indicators are calculated:

$$w_i=\frac{1-e_j}{{\Sigma }_{j=1}^n\left(1-e_j\right)}$$

(10)

The weight of each indicator under objective data w_i = (w₁, w₂, w₃, …, w_n) is obtained (

Table 11. Game theory portfolio assignment weights.

Standardized Layer	Guideline Layer Weights	Indicator Name	Subjective Weights, a	Objective Weights, w	Combined Empowerment Indicator Layer Weights, w*
Policy transmission, A	0.1062	Participation in sectoral linkages, A1	0.1134354	0.0455	0.106182601
Potential evaluation, B	0.1629	Community public service facilities improvement, B1	0.1474660	0.0547	0.137558748
		Community size suitability, B2	0.0042455	0.0626	0.010470794
		Community population size fitness, B3	0.0078458	0.0734	0.014845615
Demand of the population, C	0.1359	Resident needs survey, C1	0.1474660	0.0393	0.135917148
Front-end operations, D	0.1374	Functional space planning and design front, D1	0.1134354	0.0411	0.105714365
		Operator pre-planning participation, D2	0.0313177	0.0346	0.031663132
Space scene, E	0.0251	Degree of sharing of public service facilities, E1	0.0071325	0.0567	0.012428125
		Stock space utilization, E2	0.0094149	0.0397	0.012650397
Continuing operation, F	0.1056	Operational services perceptible, F1	0.0613828	0.0667	0.061954809
		Scene continues to open, F2	0.0438448	0.0424	0.043690491
Multi-subject, G	0.1004	Long-term stability of operating entities, G1	0.0173987	0.0416	0.01998133
		Degree of diversification of governance actors, G2	0.0050946	0.0344	0.008219458
		Residents’ satisfaction with multi-objective governance, G3	0.0736593	0.0596	0.072160543
Balance of funds, H	0.1121	Working capital closure degree, H1	0.0810253	0.0576	0.078522863
		Degree of diversification of funds, H2	0.0313177	0.0527	0.033605145
Model optimization, I	0.0284	Transportation merchants’ quality degree, I1	0.0071325	0.0558	0.012327133
		Community operations dynamics optimization, I2	0.0124277	0.0463	0.016040684
Intelligent operation, J	0.0861	Degree of digital platform intensification, J1	0.0112979	0.0437	0.014754153
		Effective operation of digital platforms, J2	0.0736593	0.0517	0.071312467

); subsequently, the expert correspondence data are used to obtain the weight of the indicators under the main and objective evaluation.

5.3. Determination of Combined Game Theory Weights Based on Subjective and Objective Hedging

As can be seen from

Table 12. Results of TOPSIS-based community rankings.

Community Name	Positive Ideal Solution Distance (D+)	Negative Ideal Solution Distance (D-)	Composite Score Index
Qibei Future Community	0.70839738	0.48741435	0.40760124
Dohuli Future Community	0.52395624	0.67227702	0.56199493
Future Community of Harmony	0.6356325	0.52383431	0.45178897
Wang Jiang Yuan Future Community	0.4870308	0.74554473	0.60486738
ZTE Future Community	0.47705928	0.67207857	0.58485461
Riverside Future Community	0.59913773	0.69445709	0.53684282

, there is a certain divergence between experts and objective data. Objective data are more inclined to give high weight to some indicators, which are influenced by objective material conditions’ data, while experts give high weight to the indicators which are more based on group preference for the links in which people are involved. Therefore, we adopt the game theory comprehensive assignment, coordinate the experts’ assignment with the entropy weighting method to offset the experts’ overly subjective judgment, and solve the problem of realistic factors that cannot be taken into account by the information entropy. The subjective weights and objective weights are regarded as the two parties involved in the game, and the combination assignment based on game theory will be a linear combination of subjective and objective weights; the optimal evaluation index weights are obtained by solving the minimization difference between different weights. The basic steps are as follows:

We assume a basic set of weight vectors, w_k = {w_k1, w_k2, w_k3, ……, w_km},k = 1, 2, ……, L, for the Future Community-wide management evaluation system, where w_k is a set of weights for the kth assignment method, and m is the number of indicators in the system.

We denote the composite weight vector as any linear combination of the L basic weight vectors, and c_k is denoted as the combination coefficients:

$$w={{\displaystyle \sum }}_{k=1}^L{c}_k{w}_k^T,c_k>0$$

(11)

The optimal linear combination of the two subjective and objective weights, with the objective of minimizing the deviation, yields the optimal weight value, w*, which is a function of $\mathrm{the}\mathrm{following}$:

$$\min \Vert w-w_k\Vert {}_2,k=1,2$$

(12)

The above equation is equivalently transformed to a linear equation with optimal first-order derivative conditions, according to the matrix differentiation property:

$$\left[\begin{array}{cc}{w}_1{w}_1^T& w_1{w}_2^T\\ w_2{w}_1^T& w_2{w}_2^T\end{array}\right]\left[\begin{array}{c}{a}_1\\ a_2\end{array}\right]=\left[\begin{array}{c}{w}_1{w}_1^T\\ w_2{w}_2^T\end{array}\right]$$

(13)

The combination coefficients a₁ and a₂ are derived for the above column matrices and then normalized to finally obtain the combined weights based on the game theory combination assignment, as follows:

$$w^{\ast }=a_1^{\ast }{w}_1^T+a_2^{\ast }{w}_2^T$$

(14)

Among them, we have $a_1^{\ast }=\frac{a_1}{a_1+a_2}, a_2^{\ast }=\frac{a_2}{a_1+a_2}$.

The computation of the combination assignment is accomplished using the relevant MATLAB code and the combination coefficients a₁* = 0.8364 and a₂* = 0.1636. The combined weights are shown in Table 11.

5.4. System Application—Sample Community Ranking Based on TOPSIS Methodology

Using the TOPSIS method, in order to ensure that all numbers are comparable and consistent both horizontally and vertically, the sample data were first normalized to keep the range between 0 and 1.

The normalization matrix is constructed as follows:

$$H_{ij}=\frac{x_{ij}}{\sqrt{{{\displaystyle \sum }}_{k=1}^n{\left(x_{ij}\right)}^2}}$$

(15)

The distance of the i^th evaluation object from the maximum value is defined, where wj is the composite weight derived from that obtained in Section 4:

$$D_i^{+}=\sqrt{{{\displaystyle \sum }}_{j=1}^m{w}_j{\left(H_j^{+}-h_{ij}\right)}^2}$$

(16)

The distance of the i^th evaluation object from the minimum value is then defined:

$$D_i^{-}=\sqrt{{{\displaystyle \sum }}_{j=1}^m{w}_j{\left(H_j^{-}-h_{ij}\right)}^2}$$

(17)

Finally, the distance of the object from the optimal solution is evaluated:

$$C_i=\frac{D_i^{-}}{D_i^{+}+D_i^{-}}$$

(18)

where D+ and D− values represent the distance (Euclidean distance) between the evaluation object and the optimal or worst solution (i.e., A+ or A−), respectively. The practical significance of these two values is that they indicate the distance between the evaluation object and the optimal or the worst solution. A larger value indicates that the distance is greater; the larger the D+ value of the research object, the greater the distance from the optimal solution, and the larger the D− value, the greater the distance from the worst solution. It is understood that the smaller D+ value, the larger the D− value.

Within the composite degree score (the Ci value), if the D− value is relatively larger, the research object is further away from the worst solution, and therefore the research object is better; a larger Ci value indicates that the research object is better.

5.5. Calculation Results and Discussion

This section offers a comprehensive assessment and ranking of the chosen communities, utilizing the TOPSIS method as an information aggregation approach. The viability of the indicator system is confirmed by comparing the community rankings based on the linear weighting of comprehensive weights, with comprehensive evaluation rankings derived via the TOPSIS method. The calculated results, obtained through SPSS, are presented in Table 12.

Following a thorough evaluation employing the TOPSIS method, this paper presents the ranked order of management processes across the six communities as follows: (1) Wangjiangyuan Future Community, (2) Zhongxing Future Community, (3) Dohuli Future Community, (4) Riverside Future Community, (5) Mutual Future Community, and (6) Qibei Future Community. The overall findings from the data survey indicate a notable level of satisfaction with the governance of all communities. This suggests that the establishment of Future Communities has demonstrably contributed to enhancing convenience and overall quality of life for residents.

The Future Community of Wangjiangyuan attained the top ranking due to its comprehensive fulfillment of numerous indicators. As one of the initially successful Future Communities in Zhejiang Province, Wangjiangyuan benefits from heightened attention from the local government, which optimizes the Future Community based on the experiences of preceding communities. It secures the leading position in the sample across six key indexes: the integration of participating departments, the optimal community population size, the thoroughness of research into resident needs, the sustained openness of the community scene, the closed-loop efficiency of operating funds, and the dynamic optimization of community operations. Remarkably, the numerical scores for the first two indicators substantially surpass those of other samples. Moreover, the Wangjiangyuan Future Community excels because of the long-term stability of its operating body, the degree of diversification of its governing body, and the degree of diversification of its funding, securing the second position among the samples in each of these categories. This exceptional performance underscores the community’s commitment to not only meeting immediate needs but also to fostering enduring stability and diversity in its governance and financial structures.

The Qibei Future Community represents the initial wave of approved Future Communities at the county level within the province. Being among the first in this category, it faces a unique challenge—the absence of relevant precedents to draw lessons from. Consequently, it occupies the last position in the comprehensive evaluation of the entire Future Communities’ system, a ranking that aligns with its distinct circumstances. As the inaugural local Future Community, Qibei must navigate the creation process without the benefit of learning from analogous cases, relying solely on its own initiatives. This self-directed approach has, however, resulted in some inevitable oversights. Primary among these challenges are the lack of robust front-end operations, inadequate operational hours for community facilities, the limited engagement of diverse stakeholders, and a monolithic approach to fund management. These challenges highlight the inherent difficulties faced by pioneering local Future Communities in forging a path without the precedent of established models.

6. Conclusions

This study addresses the practical aspects of Future Community development, focusing on the creation of an evaluation system for holistic management. It delves into the construction of an evaluation index system that accounts for the distinct challenges and features of Future Community development. A comprehensive evaluation system for Future Communities oriented towards holistic management is proposed, encompassing ten key facets: policy dissemination, potential assessment, residents’ needs, operational readiness, spatial scenarios, continuous operation, diverse stakeholders, financial equilibrium, model optimization, and intelligent operation. Within this framework, approximately 20 critical indicators were selected through the Fuzzy Delphi Method (FDM), resulting in the establishment of a scientific and rational evaluation index system. To determine the weights of specific indicators within the constructed evaluation model, a game theory-based combinatorial empowerment approach was employed. This helped to minimize the influence of individual objective and subjective factors, fostering a more balanced evaluation. Subsequently, real-world cases are subjected to comprehensive analysis and evaluation, using the TOPSIS method. The application of this evaluation system to the case validates its feasibility and practicality for whole-process management within Future Communities. This evaluation system serves as a valuable tool to promote holistic management practices in Future Communities. Through further applications, additional samples, and ongoing refinements, it has the potential to offer substantial support for decision-making and evaluation processes across various Future Community developments within the region.

It is crucial to acknowledge that the theories, methods, and models presented in this paper possess general applicability; however, caution against directly transplanting the indicator system from one country to another is advised. Owing to distinct national conditions, the application of the evaluation system outlined in this paper to countries and regions beyond China necessitates thoughtful adjustments to the specific indicator system. When implementing this evaluation system in diverse contexts, particularly outside of China, we recommend that the indicator system is adapted judiciously, building upon the preliminary framework provided in this paper. For optimal results, employing the fuzzy comprehensive decision-making method outlined in this study may prove beneficial. This involves conducting a meticulous examination of indicators’ composition and their respective weights, thereby facilitating the development of a customized and standardized system that aligns with local characteristics and conditions. This approach ensures a rigorous and context-specific evaluation process tailored to the unique circumstances of each locale.