Research on CICP

1. (1) Themes of the existing studies

Research on CICP currently spans multiple dimensions, including technology, industry, education, and management, aims to provide actionable recommendations. The existing studies can be grouped into the following thematic areas:

Policy text quantitative analysis

In recent years, the approach to analyzing policy texts has shifted from qualitative to quantitative ones, focusing on a combination of descriptive implementation and statistical validation to facilitate scientific policy formulation and decision-making [15]. As a unique type of document, policy texts have adopted and innovated upon the theoretical methods of bibliometrics from the academic literature due to their structural similarities [7]. The scientific nature of this approach lies in its use of extensive data analysis to extend the scope and coverage of research. However, based on large samples, this method also tends to overlook individual characteristics in pursuing standard textual features [7], often necessitating the integration of content analysis in the research process.

Several research perspectives exist, such as policy evolution, tools, and evaluation. A multi-tiered methodological approach [15] has been formed: (1) Utilizing data statistics to analyze characteristics such as text publication time, keywords, and issuing institutions; (2) interpreting the content of texts from aspects such as policy themes, goals, and tools, combined with qualitative analysis for a logical explanation; (3) delving deeper into policy information and its connotations for a comprehensive evaluation and recommendations. With societal development, new research methods have emerged with the rise of big data technologies, such as examining the interrelationships between policy-making entities from a network perspective and enhancing understanding of text semantics using knowledge graphs. The differences in samples, algorithms, and tools lead to varied research outcomes. An appropriate analysis framework needs to be constructed based on the attributes of the research objectives.

Research on policy evaluation

Policy evaluation is a feedback mechanism for refining existing policies and a predictive tool for shaping new ones [16], guiding their formulation, adjustment, and optimization [17]. This process requires a comprehensive review of the policy system, utilizing scientific standards and methods to judge and measure policy texts [17,18,19].

In the early times, policy evaluation was predominantly empirical. Since the 1970s, policy evaluation has evolved from empirical to normative standards, focusing on value judgments. This shift is exemplified by Suchman’s five evaluation types [20] and Poland’s “three E” framework [21]. As research progresses, policy evaluation now integrates content analysis with expert insights. It concentrated on semantic analysis, primarily through content analysis, by comparing high-frequency and thematic words in policy texts [22,23]. While qualitative evaluation has merits in analysis and summarization, its subjectivity regarding evaluation indicators and lack of objective reference points has been criticized [24].

Consequently, there is a growing trend of incorporating quantitative approaches in policy evaluation. Beginning as early as 1978, Libecap [25] compared current and past policies to create a legal change index for evaluating Nevada’s mining laws by sorting out the scores of all policies throughout the year. After that, scholars worldwide have researched quantitative analysis in policy evaluation, most relying on scoring, and the topics cover finance, science, technology, healthcare, and the environment. [23,26–28]

In contemporary policy evaluation, the focus shifts from exclusively qualitative or quantitative methods to integrating both [29], blending qualitative approaches like case studies and expert reviews with quantitative analyses using mathematical models. Empirical approaches such as text data mining [30,31], social network analysis [32], and fuzzy comprehensive evaluation [33] have become comprehensive. Standard empirical tools include PSM-DID model analysis [34], instrumental variables [35], and synthetic control [36]. Advancements in quantitative analysis tools for policy texts also show new progress. Natural language processing (NLP) tools convert text into numerical data, enabling the analysis of semantic information [37,38]and extraction of high-frequency vocabulary, thus distilling key policy messages [39]. This approach, now prevalent in policy text analysis [40–43], typically employs software like Python [44], Rost Content Mining 6.0 [45], and ROSTCM 6.0 [46,47]. Furthermore, the PMC Index model has been introduced to evaluate individual policies and facilitate comparative analysis across multiple policies, marking a significant stride in policy evaluation.

Application of the PMC index model

The PMC index model was established by Ruiz Estrada et al. [8,48]and is based on the Omnia Mobilis hypothesis [49] This hypothesis suggests that everything is interconnected and dynamic, urging a comprehensive approach to assessing the consistency of policy texts. The PMC index model merges Cartesian spatial application with text mining, offering a multi-dimensional evaluation that visualizes policy texts’ strengths and weaknesses. Scholars have applied it to evaluate digital transformation policies [12], land protection policies [50], big data development policies [51], insurance policies [52,53], pork industry policies [46], disaster relief policies [54], ecological protection compensation policies [47], and emergency response policies [38] evaluation. Zhang et al. [55–59] applied it to innovation policies, financial policies, new energy vehicle subsidies policies, and more, contributing significantly to China’s policy evaluation framework. Furthermore, based on the PMC index model, Zang et al. [60] quantitatively analyzed China’s AI policies and proposed policy formulation suggestions in line with research frontiers, and Hu et al. [61] quantitatively evaluated China’s robot industry policies, providing a decision-making basis for policy optimization. These applications demonstrate the feasibility and value of using the PMC index model for evaluating CICP.

Contributions of this research

Given the lack of application of quantitative analysis methods in CICP, the PMC Index model is introduced into the analysis process in this research, and a multi-dimensional discussion of the policy features and formulation is conducted. Furthermore, considering the policy’s timely and ever-evolving nature, this research integrates a dynamic analysis of temporal changes to compare the CICP at various phases. The approach facilitates a more comprehensive understanding of the CICP’s evolution and status features. As a result, it allows for a more nuanced exploration of policy trends. The systematic and dynamic methodology in this research offers a fresh perspective for policy analysis and evaluation, moving beyond conventional static methods.

Methodological framework

Based on references [13,14,62–64], this research constructed an advanced multi-dimensional analytical framework that applies text mining and data visualization analysis techniques to rigorously analysis the CICP. The framework covered evaluation, discussion, and optimization recommendations for CICP, and be executed according to five parts of advanced work (see Fig 1), including the policy evaluation based on the PMC Index Model and the dynamic analysis of policy evolution, status, and trend, offering data-driven recommendations for the optimization of CICP.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Analytical framework.

https://doi.org/10.1371/journal.pone.0326505.g001

Research method and work procedure

This study employed methods such as text encoding, data analysis, and graphical visualization. The research process consisted of five stages: data acquisition, data processing, computation and analysis, results discussion, and conclusion. (see Fig 2) After the computation and visualization of the data, the five parts of advanced work constructed in the analysis framework above (see Fig 1) were carried out, and the corresponding findings and prospects were proposed. The following is an introduction to the research methods and working procedures according to contents established by the framework, these are also described in detail in Fig 2.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Research method and work procedure.

https://doi.org/10.1371/journal.pone.0326505.g002

1. (1) Evaluate the formulation of CICP based on the PMC index model

Firstly, the relevant texts of CICP were collected and collated based on the information disclosure platform, and a research database was established. Then, the PMC index model was used to scientifically analyze and evaluate the formulation of CICP, and then the phases of CICP’s development were summarized and discussed. In this process, the evaluation of CICP using the PMC Index Model constitutes the most critical section. As delineated in the preceding literature review, the PMC index model enables quantitative assessment of policy comprehensiveness and coherence through the establishment and scoring of multidimensional indicators. This method facilitates the identification of strengths and weaknesses in policy formulation through data-driven analysis, thereby providing empirically grounded insights for future policy optimization. The subsequent section will elaborate on the application methodology of this model with specific exemplifications.

The evaluation procedure follows a standard four-step process for calculating the PMC Index [8], combined with content analysis of policy texts. The steps are outlined as follows:

Step 1: Define computational variables and identify parameters

This step involves setting up primary and secondary variables based on the selected policies within the sample. Each variable is assigned specific parameters, which are configured to align with the text content and intended policy goals. In general practice, primary variables are employed to assess the consistency of policies. These variables exhibit minimal variation in their configuration across different studies, allowing researchers to reference classical settings from existing applications of this model. By referring to the classic research methodology of authoritative scholars and combining it with the development demands of China’s intelligent construction, the selection and setting of primary variables can be carried out. Secondary variables, operating within the scope of primary variables, facilitate a more granular analysis and comparison of various dimensions and factors in policy development. These variables are particularly instrumental in identifying deficiencies and shortcomings in policy formulation. Consequently, their determination requires customization based on the specific characteristics of the analytical subject. The selection of corresponding variables typically derives from a systematic synthesis of research findings within the subject’s domain, complemented by empirical feedback from practical implementation experiences.

For example: the policy text (e.g., ‘Policy P0’) is evaluated within a standardized framework of ten primary variables, including “X1 Policy Nature,” “X2 Policy Effectiveness,” and so forth. Within the scope of these primary variables, a hierarchical system of secondary variables is systematically developed. To illustrate, X1 incorporates two secondary variables: “X1:1 Prediction” and “X1:2 Regulation.” Similarly, X2 is structured with three secondary variables: “X2:1 Short-term,” “X2:2 Medium-term,” and “X2:3 Long-term,” among others. It is noteworthy that certain primary variables, such as X10, may remain without secondary variables based on analytical requirements. (see Table 1)

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Evaluation index system and criteria for P0.

https://doi.org/10.1371/journal.pone.0326505.t001

Step 2: Conduct calculations using the PMC index formula

In this step, calculations are carried out using the classical PMC Index formula. Through a comprehensive content analysis of the policy text, the secondary variables of the target policy can be quantified by referencing Formula (1) and Formula (2). This is done by evaluating the presence or absence of specific variables in the policy texts and assigning binary values (see Table 1).

(1)

(2)

Secondary variables are treated in binary form, with equal weighting applied. Once secondary variables are processed, primary variables are calculated using Formula (3), and the overall PMC Index is computed using Formula (4). The aggregate score of a policy formulation is derived through the summation of ratings across all primary variables.

(3)

(4)

(i = primary variable; j = secondary variable; n is the total number of secondary variables.)

For example Generally, a multi-input-output table is used to handle extensive data for each variable, serving as the foundation for analyzing primary variables. Binary values are assigned to secondary variables, and these are used to calculate the primary variables (see Table 2). For instance, during the textual analysis of Policy P0, the scoring process was conducted as follows: Regarding the “X1 Policy Scope” variable, we classified the policy scope as “X1:1 Prediction” rather than “X1:2 Regulation,” resulting in a score of 0.50 for X1. For the “X2 Policy Effectiveness” variable, we determined that the policy’s efficacy primarily targets short-term and medium-term outcomes rather than long-term effects, yielding a score of 0.67 for X2. This scoring methodology was consistently applied across subsequent variables. Specifically, for X10 “Policy Disclosure,” we assessed the policy as fully transparent, assigning a maximum score of 1.00. Through the aggregation of scores from all ten primary variables, we obtained a composite score of 7.50, which represents the final evaluation of P0 policy formulation based on the PMC index model.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Multi-input–output table for P0 (for example).

https://doi.org/10.1371/journal.pone.0326505.t002

Step 3: Policy rating based on PMC index

The composite score can be subsequently categorized according to predefined grading standards, facilitating both hierarchical classification and comparative analysis among different policies. The rating criteria are set as follows in normal: [0, 5.99] is Average, [6, 6.99] is Acceptable, [7, 7.99] is Positive, [8, 8.99] is Excellent, and [9,10] is Perfect. Such an analytical framework also enables the identification of relative strengths and weaknesses, thereby providing a robust basis for policy evaluation and improvement.

For example Based on the computational results derived from the PMC model, Policy P0 attained a score of 7.50, which, according to the established evaluation criteria, categorizes its formulation quality within the “Positive” classification tier. This methodological framework can be systematically replicated to evaluate and score other policies, thereby enabling comparative analysis and cross-policy benchmarking.

Step 4: Construct PMC surfaces based on PMC index

PMC surface drawing is employed to facilitate multi-dimensional policy evaluation and allow for a deeper single-indicator analysis. This method is especially useful for assessing the internal consistency of policies and comparing various approaches. The PMC surfaces are calculated using third-order matrices as Formula (5), which provide a visual depiction of the evaluation outcomes. The variance between the evaluated policy and an ideal “perfect” policy is highlighted, which can be quantified using the concavity index [59]. The concavity index is calculated as the perfect score of 10 minus the variable’s PMC evaluation score, and the concavity index composition refers to the proportion of a variable’s concavity index relative to the total concavity indices across all variables, thereby enabling the quantification of policy formulation deficiencies within specific variable dimensions.

(5)

For example Based on the evaluation scores of P0’s nine primary variables (note that X10 is exclusively designated for assessing policy transparency and thus excluded from this analytical phase), a three-dimensional matrix is constructed as specified in Formula (5). This matrix serves as the computational foundation for generating corresponding surface diagrams using analytical software such as EXCEL. Through integrated analysis combining graphical representations with the composition of primary variables’ concavity index (see Fig 3), we can systematically identify and quantify the shortcomings in the policy formulation process of P0. The graphical analysis reveals that Policy P0 demonstrates optimal formulation in the dimensions of X3, X4, and X7, as evidenced by their respective scores and graphical representations. However, deficiencies are identified in the X1, X6, and X9 dimensions, indicating areas requiring substantial improvement and refinement in the policy formulation process.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. PMC Surfaces and Concavity Index Distribution for P0 (example).

https://doi.org/10.1371/journal.pone.0326505.g003

1. (2) Discuss the evolution of CICP

Following the quantitative evaluation of CICP’s formulation quality and its developmental stage classification, a comprehensive investigation into the distinctive features of CICP’s evolution is warranted. This subsequent analysis aims to delve deeper into the intrinsic patterns and unique attributes that have shaped CICP’s developmental trajectory, thereby providing valuable insights for future policy refinement and strategic planning. The dimensions of analysis include policy themes, policy types, and intergovernmental relations, focusing on how policy supply and demand drive these changes examining government decision-making and governance shifts. The NLP tool, ROSTCM6 from Wuhan University, provides text analysis methods, including word segmentation and frequency assessment. It helps extract keywords and create semantic networks, revealing the main themes and hotpots in policy discussions. Additionally, the research assesses policy effectiveness and government collaboration using policy types and issuing institutions’ statistics.

1. (3) Discuss the status of CICP

Having identified the evolutional features of CICP, the analysis proceeds to discuss the characteristics of policy status. The dimensions of analysis include policy goals and tools. Policy tools are interventions policymakers adopt to achieve their goals [65], reflecting the government’s policy values and governance concepts [66]. The research considers the structural characteristics of policies and government intervention methods by analyzing the distribution of policy goals and tools in different periods and the coupled application of goals and tools.

1. (4) Identify the trend and find the gaps in CICP

This research utilizes the CNKI database for a three-year literature review, employing the software VOSviewer for bibliometric analysis. This method generates a table of high-frequency keyword clusters essential for identifying key trends in China’s intelligent construction field. The aim is to find the policy gaps and discover potential optimization areas of the CICP by integrating the findings of policy evaluation based on the construction of PMC Surfaces.

1. (5) Propose the suggestions of CICP

Based on the findings and discussions above, the paper advances policy recommendations to foster the rapid progression of intelligent construction. These suggestions enhance the comprehensiveness and adaptability of policy formulation in this domain.

Data sources and sample description

Research sample processing and coding.

This research gathered 30 national policy texts on intelligent construction, adhering to the outlined search process and selection principles. Table 3 presents these texts, coded for statistical analysis. Table 4 demonstrates the encoding rules developed based on the structural characteristics of the policy texts, including further encoding and categorizing the content of the text clauses.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Research sample of national intelligent construction policy texts (2010-2022).

https://doi.org/10.1371/journal.pone.0326505.t003

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Example of encoding rules for text clauses.

https://doi.org/10.1371/journal.pone.0326505.t004

Policy evaluation based on the PMC index model

Parameter setting.

As delineated in the preceding methodological exposition, the selection of primary variables in this investigation is grounded in established classical configurations derived from extant research. The formulation of secondary variables, conversely, is predicated upon a nuanced consideration of the specific developmental status and exigencies inherent in intelligent construction policies, thereby ensuring the analytical framework’s responsiveness to contemporary policy dynamics.

Based on Estrada’s theory [48] and incorporating insights from other scholars [57,59,60] in policy evaluation, this study selected ten primary variables to analyze thirty policies within the feature analysis of CICP. The thirty-six secondary variables were carefully selected and configured within the framework of primary variables, incorporating a comprehensive review of intelligent construction policy research in China, complemented by expert consultations and substantiated by practical implementation experiences [3,9–14]. This rigorous selection process ensures the variables’ relevance and applicability to the specific context of the CICP landscape. The specific variables and their corresponding parameters are detailed in Table 5, which outlines the evaluation criteria used for the analysis.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Evaluation Index System and Criteria for CICP.

https://doi.org/10.1371/journal.pone.0326505.t005

“Policy Scope” serves as a fundamental dimension, examining the various forms policies take in promoting intelligent construction development. This includes predictive functions that anticipate industry trends, regulatory mechanisms that establish operational frameworks, advisory roles that provide guidance, supportive measures that facilitate implementation, and guiding principles that shape strategic direction, along with provisions for unique policy characteristics. “Policy Effectiveness” also serves as a crucial dimension, distinguishing between short-term, medium-term, and long-term impacts. This temporal analysis enables the assessment of policy sustainability and phased outcomes, providing insights into the dynamic nature of intelligent construction development.

To ensure comprehensive sectoral coverage, the “Policy Field” addresses the primary areas influenced by CICP, encompassing economic, social, technological, and environmental considerations. This broad perspective is complemented by the “Policy Perspective” analysis, which examines policies across macro, meso, and micro levels of governance, facilitating a thorough understanding of policy impacts at different operational scales.

The evaluation framework further investigates policy implementation through two critical dimensions: “Policy Function” and “Incentives and Constraints”. “Policy Function” focuses on the operational mechanisms employed to promote intelligent construction, including government procurement processes, standardization guidance, regulatory frameworks, technological innovation support, and industry development strategies. The “Incentives and Constraints” evaluates the specific policy instruments utilized, such as investment subsidies, legal protections, talent incentives, tax benefits, intellectual property rights, and support services.

To assess the quality of policy formulation, the “Policy Evaluation” establishes rigorous criteria, examining aspects such as research adequacy, objective clarity, planning detail, and scientific rigor in scheme design. This quality assessment is further enhanced by examining the “Issuing Institution”, which analyzes the hierarchical level of policy-making bodies, providing insights into the authority and scope of policy decisions. The framework also incorporates “Policy Recipients” as a critical dimension, analyzing the implementation scope and effectiveness through the examination of target agencies and their operational levels.

Finally, the “Policy Disclosure” assesses the public accessibility of policies and their potential for broad societal impact, completing the comprehensive evaluation framework.

This integrated approach enables a systematic analysis of CICP, ensuring robust evaluation of their formulation quality, implementation mechanisms, and potential impacts across various sectors and governance levels while maintaining the flexibility to adapt to the evolving nature of intelligent construction development.

Building the multi-input–output tables and setting parameter values.

This research involves calculations with previously defined primary variables, constructing multi-input-output tables for the thirty policies (see Table 6), and integrating these with the standard PMC index formula. The values for secondary variables (Table 5), are assigned based on the policy texts’ content analysis, following the methodology outlined in Formulas (1) to (2). All secondary variables are considered equally important and are thus weighted the same. When a policy encompasses a particular secondary variable, its parameter value is set to 1; otherwise, it is set to 0.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Multi-input–output table.

https://doi.org/10.1371/journal.pone.0326505.t006

Calculating and detailed description of the PMC index.

Each primary variable is calculated using Formula (3), and subsequently, the PMC index is determined through Formula (4) based on these values. The 30 policy texts are then rated according to the PMC index, as the rating criteria outlined in the methodology section. Then, the PMC indices and corresponding ratings for the 30 policy texts are derived, as indicated in Table 7. The findings indicate:

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Calculation and evaluation results of PMC index for 30 policy texts.

https://doi.org/10.1371/journal.pone.0326505.t007

1. (1) The average PMC index for 30 policies stands at 6.93, classified within the “Acceptable” at the policy evaluation level, closely approaching the “Positive” benchmark. The result suggests that CICP are well-rounded, addressing various dimensions and establishing a solid foundation for effective implementation.
2. (2) Each of the 30 policies has a PMC index exceeding 5.00, indicating the absence of poorly formulated policies.
3. (3) Within the quality evaluations of these 30 policies, two achieved “Excellent” (7%), fourteen “Positive” (46%), eight “Acceptable” (27%), and six “Average” (20%). These outcomes demonstrate that the formulation quality of 80% of the policies is considered acceptable, with over 50% having a constructive impact, reflecting well on the general formulate quality of CICP.
4. (4) Among these policies, “Decision on Accelerating the Cultivation and Development of Strategic Emerging Industries”(P1) [67] introduced by the State Council in 2010 marks the inception of the CICP framework, bearing a PMC index of 7.12 and classified as “Positive.” The policy signifies its beneficial guidance on China’s intelligent construction evolution.
5. (5) The highest quality of policy formulation was observed in the “Guiding Opinions on Promoting the Coordinated Development of Intelligent Construction and Construction Industrialization” (P15) [68], issued in 2020 by the MHURD along with 13 other departments, achieving a PMC index of 8.36 and classified as “Excellent.” The policy indicates its significant role in fostering the rapid advancement of intelligent construction in China.

Classification and definition of policy phases.

Statistics based on the data in Table 7, a comparative analysis of policy evaluations across different periods reveals that the introduction of policy “P15” in 2020 marks a pivotal milestone in the evolution of CICP. As illustrated in Fig 4, the analysis further reveals a significant shift in the development of CICP occurred around the year 2020.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Comparison of policy ratings in different periods.

https://doi.org/10.1371/journal.pone.0326505.g004

The pre-“P15” period (2010–2019) was characterized by a gradual development pace, during which the government issued 14 intelligent construction-related policies, averaging 1.4 per year. The policies formulated in this phase had PMC index values ranging from 5.57 to 7.65, with 50% classified as “Acceptable” and 36% as “Positive,” but none achieving “Excellent” ratings.

In contrast, the post-“P15” period (2020–2022) witnessed a substantial acceleration in policy development, with the government releasing 16 policies within three years, averaging 5.3 policies annually. These policies demonstrated improved quality, with PMC index values ranging from 5.24 to 8.36, including 6% rated as “Acceptable,” 56% as “Positive,” and 13% as “Excellent.”

This comparative analysis indicates a transformative paradigm shift in CICP development following the introduction of Policy “P15”, manifesting in both quantitative and qualitative dimensions. The substantial increase in policy issuance frequency (from 1.4 to 5.3 policies per year) and the emergence of “Excellent” rated policies (from 0% to 13%) collectively demonstrate enhanced policy-making capacity and strategic focus.

The empirical analysis above substantiates that CICP demonstrates distinct phased development characteristics with identifiable transition points. Building upon data-driven evidence, our comprehensive content analysis of 30 CICP texts further validates this developmental trajectory, revealing a systematic progression from the Cultivation and Exploration Phase (CEP: 2010–2019) to the Development and Promotion Phase (DPP: 2020–2022), as delineated in Fig 5.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Policy phases.

https://doi.org/10.1371/journal.pone.0326505.g005

During the foundational CEP period, the Chinese government established critical policy frameworks through landmark documents such as the “Decision on Accelerating the Cultivation and Development of Strategic Emerging Industries” (P1), “Outline for the Development of Information Technology in the Construction Industry (2011-2015)” (P3), “Made in China 2025” (P6), and “Guidance on Vigorously Developing Prefabricated Buildings”(P9). These initiatives primarily focused on technological infrastructure development, particularly promoting digital technologies including BIM, 3D printing, and IoT in prefabricated construction. Through national-level innovation support and industrial guidance, CEP facilitated the construction industry’s initial transition toward informatization and digital transformation. However, the phase was characterized by restricted policy scope, practical application challenges, and noticeable stagnation periods during 2012-2013 and 2018-2019, reflecting its exploratory nature and limited implementation breadth.

The subsequent DPP represents a transformative phase initiated by the 2020 policy “Guiding Opinions on Promoting the Coordinated Development of Intelligent Construction and Construction Industrialization” (P15), which introduced comprehensive strategies emphasizing construction industrialization, digital transformation, and intelligent system integration. This phase reached a significant milestone with the 2022 “14th Five-Year Plan for Construction Industry Development” (P25) [69], establishing ambitious objectives for global leadership in intelligent construction. Characterized by extended temporal efficacy, expanded application domains, clearer strategic objectives, diversified implementation tools, and enhanced synergy with construction industrialization, DPP policies have significantly accelerated intelligent construction adoption across China’s building sector.

This evolution from CEP to DPP reflects a fundamental shift from foundational policy frameworks to comprehensive implementation strategies, demonstrating CICP’s adaptive capacity in responding to technological advancements and industry demands. The phase transition marks a critical development in the CICP landscape, moving from exploratory experimentation to rapid development and maturation, while establishing a robust foundation for future policy innovation and industry transformation.

Discussion on phase shift.

Table 8 presents the comparative PMC index analysis across two phases. The CEP yielded an average score of 6.82 (“Acceptable”), while DPP achieved 7.03 (“Positive”), demonstrating enhanced policy quality. Fig 6 reveals significant rating distribution shifts: “Average” policies increased from 2 to 4, “Acceptable” decreased from 7 to 1, “Positive” rose from 5 to 9, and “Excellent” emerged from 0 to 2. These improvements indicate DPP’s stronger impact in advancing China’s intelligent construction. At the same time, China’s intelligent construction market has also risen rapidly from 150 billion yuan in 2019–250 billion yuan in 2022, with an average annual growth rate of more than 20%. This is further evidence of the effectiveness of the DPP policy.

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. The average PMC index values of the policies in two phases.

https://doi.org/10.1371/journal.pone.0326505.t008

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Comparison of policy ratings between CEP and DPP.

https://doi.org/10.1371/journal.pone.0326505.g006

The comprehensive content analysis of CICP texts also reveals a significant evolution in policy formulation approaches, transitioning from the exploratory and tentative strategies of the initial phase (CEP) to more systematic, comprehensive, and sustainable frameworks in DPP, with stronger integration of practical applications. This transformative shift holds profound implications for CICP development, as evidenced by several critical dimensions of policy improvement.

Firstly, the transition to a more practical policy approach represents a fundamental paradigm shift from experimental technology promotion to mainstream application of intelligent construction principles. The introduction of “P15”, with its emphasis on construction industrialization and digitalization, has profoundly impacted construction practices by encouraging widespread adoption of new technologies and implementation of standardized processes across the industry. This shift is particularly evident in the increased utilization of BIM technologies and prefabricated construction methods, which have become more prevalent since the implementation of DPP policies.

Secondly, the DPP’s focus on long-term goals and sustainable growth, exemplified by Policy “P25”, directly addresses the limitations observed in CEP, where policy stagnation and fragmented approaches had previously slowed progress. By establishing a comprehensive, forward-looking agenda, the DPP ensures greater scalability and sustainability of CICP. This is achieved through the introduction of financial and tax incentives, along with targeted support for Small and Medium Enterprises (SMEs), which effectively address the barriers that hindered the broader adoption of intelligent construction technologies during the first phase.

In addition, the shift from CEP to DPP was driven by several key factors, including the maturation of core intelligent construction technologies, accumulated implementation experience from pilot projects, and the construction industry’s increasing capacity for digital transformation. These developments enabled the formulation of more effective policies that better align with industry needs and technological capabilities. Especially, the DPP policies demonstrate enhanced effectiveness through their holistic nature, integrating economic, environmental, and social objectives while maintaining clear implementation guidelines and support mechanisms. This comprehensive approach has resulted in policies that are not only more frequent and higher in quality but also more responsive to the evolving needs of China’s construction industry.

However, the absence of any policy receiving a “perfect” rating indicates room for further optimization in CICP. Before offering targeted recommendations for enhancement, in-depth reviews and discussions of the evolution, status, and trends of the CICP are imperative. These discussions will integrate the historical context of policy evolution and its characteristic patterns, critically analyzing the factors contributing to DPP’s superior effectiveness compared to CEP, while identifying deficiencies and potential improvement areas within current policies. The analysis in detail will employ a dynamic analytical framework to undertake a comparative examination of policy features within the CEP and DPP across various dimensions, including types of policies, entities responsible for policy formulation, and the goals and tools of these policies. These analytical dimensions maintain conceptual consistency with the evaluation variables established in the PMC model: text type analysis corresponds to “Policy Scope”, intergovernmental relationship examination aligns with “Issuing Institution” and “Policy Recipients”, content theme analysis relates to “Policy Fields”, and objective-tool analysis connects with “Policy Function” along with “Incentives and Constraints”, and so on. By leveraging the analytical outcomes derived from the PMC model, this methodical examination will culminate in scientifically grounded recommendations for policy optimization.

Discussions of policy evolution and status

Types of policy text.

Policy texts vary in effectiveness, with “notifications” and “opinions” offering broad applicability and timeliness, significantly influencing recipients’ actions. Conversely, “guidelines,” “plans,” and “manuals” are more operational, guiding practical work. This research statistically analyzes the distribution of 30 policy texts (Table 9), comparing policies’ binding and operational nature across different phases (Fig 7). The results show a predominance of “opinions,” “notifications,” “plans,” and “outlines,” indicating the government’s promotional solid intent. However, the operational aspect of policies requires further improvement. From the CEP to DPP, the share of policies with strong constraints increased from 77% to 88%, while those with solid operational characteristics rose from 38% to 62%. This shift is primarily due to “notifications,” which act as action guides, replacing “opinions,” known for suggestion-based guidance. This transition suggests a more precise government policy direction, focusing on optimizing institutional measures and detailed tasks to enhance policy effectiveness.

Download:

• PNG
  larger image
• TIFF
  original image

Table 9. Distribution of text types.

https://doi.org/10.1371/journal.pone.0326505.t009

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Statics of text types.

https://doi.org/10.1371/journal.pone.0326505.g007

Department issuing policy.

Policy formulation involves interactions among government departments as critical actors. Analyzing their co-authorship network in issuing policies uncovers patterns of cooperation and conflict [15]. The results show (Fig 8):

Download:

• PNG
  larger image
• TIFF
  original image

Fig 8. Analysis of intergovernmental relations.

https://doi.org/10.1371/journal.pone.0326505.g008

1. (1) Inter-departmental collaboration in policy formulation is intensifying. Of the policies analyzed, 70% were issued by individual departments, notably the SC and the MOHURD. Meanwhile, 30% involved multi-departmental cooperation, especially during the DPP. This trend highlights the need for varied expertise in integrating intelligent construction with industrialization.
2. (2) Different phases show varied departmental involvement. During CEP, the SC led the initiative, focusing on technological innovation with contributions from MOHURD, the Ministry of Industry and Information Technology (MIIT), the Ministry of Science and Technology (MST), and the National Development and Reform Commission (NDRC). The participation was broader in DPP.
3. (3) Different departments have varying degrees of collaboration. MOHURD, as the central leading department, collaborates closely with MIIT, NDRC, the Ministry of Ecology and Environment (MEE), and the China Banking and Insurance Regulatory Commission (CBIRC). This phenomenon demonstrates that industrial, technological, ecological, and economic support are significant principles in government policy formulation.

Themes of policy

1. (1) High-frequency words statistics

Analyzing high-frequency thematic words in policy texts helps focus on core themes. Word distribution in the title indicates policy formulation’s macro-direction, while content word distribution reflects key implementation points. Using ROSTCM6 software, 30 policy texts from Table 1 underwent word segmentation. High-frequency words were extracted, showing frequency distribution (see Tables 10 and 11). The word frequency of titles suggests those policies primarily aimed at advancing intelligent construction in the industry. During the CEP, the focus was on informatization, shifting to green and intelligent construction in the DPP. The word frequency of content highlights technological innovation to promote intelligent construction. Words like “standard, “ “management, “ “service, “ “system, “ “mechanism, “ “model, “ “platform, “ and “data” are prominent, aligning with Chinese experts’ and scholars’ policy recommendations [3,9–11].

Download:

• PNG
  larger image
• TIFF
  original image

Table 10. High-frequency words statistics in titles.

https://doi.org/10.1371/journal.pone.0326505.t010

Download:

• PNG
  larger image
• TIFF
  original image

Table 11. High-frequency words statistics in contents.

https://doi.org/10.1371/journal.pone.0326505.t011

1. (2) Semantic network analysis

This research explores the associative structures in CEP and DPP policy texts using ROSTCM6’s semantic network analysis and Table 11’s high-frequency word statistics. The resulting concentric semantic network diagram (Fig 9) uses nodes to represent high-frequency words, with their connections indicating co-occurrence frequency. The more links a node has to other nodes, the more significant it is. During CEP, the policies focused on constructing infrastructure, improving management systems, strengthening technological innovation, and standard setting. In DPP, policies are committed to accelerating the engineering application of intelligent technologies and promoting green and low-carbon construction.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 9. Capacitive analysis of high-frequency words in text content between CEP and DPP.

https://doi.org/10.1371/journal.pone.0326505.g009

Policy goals and tools

1. (1) Policy goals

This research classifies the goals of CICP into seven categories, concluding objectives and tasks at the technology, construction, industry, service, and application level. (Fig 10) Following Table 2’s approach, this research coded 30 policies’ content clauses and quantified the policy clause provisions per goal based on the content analysis. (Table 12) The research process reveals policy shifts focus across phases (Fig 11). Key findings show consistent emphasis on “Enhancing technological innovation capability” (G2), “Improving digitalization, informatization, and intelligence level” (G3), and “Revolutionizing service and supervisory systems” (G7). The policy evolution highlights an initial focus on digitalizing construction and technology innovation in the CEP, shifting in the DPP to prioritize” accelerating the industrialization of construction” (G1), with increased attention on “promoting green construction” (G4), “ establishing industrial systems” (G5), and “expanding application scenarios “ (G6). This shift towards a more comprehensive and specific policy approach is evident, particularly in industrial and environmental applications.

Download:

• PNG
  larger image
• TIFF
  original image

Table 12. Statics of policy goal clauses in the two phases.

https://doi.org/10.1371/journal.pone.0326505.t012

Download:

• PNG
  larger image
• TIFF
  original image

Fig 10. Summary of policy goals.

https://doi.org/10.1371/journal.pone.0326505.g010

Download:

• PNG
  larger image
• TIFF
  original image

Fig 11. The proportional distribution of policy goals in the two phases.

https://doi.org/10.1371/journal.pone.0326505.g011

1. (2) Policy tools

The classification of policy tools influences the achievement of policy goals. Considering the interpretation of CICP and referencing research [13,14,60], this research categorizes the tools of CICP into three main types and fifteen sub-tools, following Rothwell and Zegveld’s model [70] of Supply-side (SP), Demand-side (DP), and Environmental Policy (EP) Tools. (Fig 12) SP Tools enhances intelligent construction by providing financial support, IT, talent development, and regulatory services. DP Tools promotes it through collaborative projects, pilot demonstrations, and enterprise support. EP Tools indirectly assist via strategic planning, legal frameworks, and financial incentives, fostering a conducive environment for intelligent construction growth.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 12. Categorization of policy tools.

https://doi.org/10.1371/journal.pone.0326505.g012

The research merged the coded content clauses of 30 policies with the previously discussed classification of policy tools for intelligent construction, compiling their distribution in various phases. (Table 13) Subsequently, the proportions of policy tools in CEP and DPP were calculated and analyzed. (Fig 13) Policy tools’ distribution highlights SP and EP tools’ dominance at 39% and 37%, with DP tools at 24%, indicating the need for more intelligent construction promotion efforts. Usage among the 15 sub-tools varies. SP tools focus on system optimization (T1-5) and research infrastructure (T1-2, T1-4), while funding and talent development (T1-1, T1-3) require support. DP tools emphasize industry applications (T2-3), with minimal use of market mechanisms (T2-2, T2-5). EP tools, led by T3-3, reflect China’s role in setting standards, suggesting an increase in institutional and financial development policy support. The shift from CEP to DPP indicates a strategic government change. SP tools show increased research (T1-2) but reduced talent development (T1-3). DP tools maintain a market approach (T2-1, T2-3, T2-4), while EP tools’ changes (T3-3, T3-5) focus on practical policy implementation for intelligent construction.

Download:

• PNG
  larger image
• TIFF
  original image

Table 13. Staticst of policy tool clauses in the two phases.

https://doi.org/10.1371/journal.pone.0326505.t013

Download:

• PNG
  larger image
• TIFF
  original image

Fig 13. Analysis of policy tool proportions in the two phases.

https://doi.org/10.1371/journal.pone.0326505.g013

1. (3) Coupling analysis of policy goals and tools

A matrix model coupling three primary policy tools, fifteen sub-tools, and seven goals illustrates the evolution of government strategies across two phases. (Fig 14) Transitioning from CEP to DPP, policy integration with primary tools has increased significantly.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 14. Coupling analysis of policy goals and tools based on the statics of clauses.

https://doi.org/10.1371/journal.pone.0326505.g014

SP tools grew from 33.33% to 43.48%, DP tools from 20.37% to 31.62%, while EP tools fell from 46.30% to 24.90%, suggesting a shift towards comprehensive development. In CEP, EP tools led, focusing on resource optimization and institutional mechanisms. SP tools like T1-4 and T1-5 supported goals like G7, while significant DP tools, like T2-3, aided multiple goals. SP tools gained prominence in DPP, particularly T1-2, aiding G3 and G2, while DP tools like T2-3 and T2-4 supported various goals. The most used EP tool was T3-3, helping goals like G4. This coupling analysis shows a strategic evolution in China’s intelligent construction, initially emphasizing EP tools for a conducive environment, then transitioning to SP and DP tools for balanced, comprehensive development.

Discussions of policy gaps and development trends

Policy gaps identification by constructing PMC surfaces.

Using PMC index values from Table 7, this research could create a PMC surface for each policy based on Formula (5), offering a more detailed visual analysis. Table 5 identifies four high-rated policies that do not reach the “Perfect” level. P15 and P25 are rated as “Excellent” with 8.36 and 8.27, respectively, while P27 and P17 achieve scores of 7.82 and 7.86.

PMC surfaces for these policies (Fig 15) highlight specific areas impacting their ratings, revealing gaps against the “Perfect” policy based on concavity index distribution. P15, guiding intelligent construction and construction industrialization, is comprehensive but needs refinement in “Policy Scope, “ “Policy Perspective, “ and “Issuing Institution. “ P17, emphasizing high-quality growth in construction industrialization with intelligent construction as a driver, could improve in “Policy Scope, “ “Policy Field, “ “Policy Perspective, “ and “Issuing Institution. “ P25, focusing on carbon peak goals and reducing the construction industry’s carbon footprint, suggests enhancements in “Policy Scope, “ “Policy Perspective, “ and “Issuing Institution. “ P27, a strategic plan for the construction industry, highlights intelligent construction as a transformational focus but requires improvement in “Policy Scope, “ “Policy Effectiveness, “ “Incentive Constraints, “ and “Issuing Institution. “

Download:

• PNG
  larger image
• TIFF
  original image

Fig 15. PMC surfaces and concavity index distribution for four policies.

https://doi.org/10.1371/journal.pone.0326505.g015

However, the analysis also indicates that the authority of the issuing institution(X8) impacts the ratings of the four policies. Given that policy formulation is led by SC and MHURD, with coordinated efforts across multiple ministries, there is already substantial authority within the single industry. Thus, this factor might not primarily influence the achievement of a perfect CICP formulation.

The findings above indicate that to achieve near-optimal policy formulation quality, CICP requires further refinements across four critical dimensions: expansion of policy scope, enhancement of policy effectiveness, broadening of regulatory perspectives, and diversification of incentive mechanisms. To establish actionable optimization guidelines and ensure the temporal relevance and sustainability of future policies, it is imperative to align CICP’s developmental trajectory with cutting-edge research frontiers in China’s intelligent construction field.

Development trends discussion with the research frontier.

This research offers integrating research frontier insights for a deeper analysis. A thorough search for “intelligent construction” was conducted in the CNKI, China’s preeminent academic literature database. After data cleaning and selection, the research analyzed seven hundred ninety-two literature samples (2021–2023). The keyword co-occurrence network was created using Vosviewer, a bibliometric tool, revealing research frontiers and hotspots in China’s intelligent construction field (see Table 14). The results show that China’s intelligent construction sector is undergoing a transformative phase driven by technological convergence and policy innovation. At its core, the integration of BIM (Building Information Modeling, FQ = 165) with artificial intelligence (FQ = 32) and digital twins (FQ = 29) has established a technological triad that reshapes traditional construction paradigms. This synergy fuels advancements in smart construction sites (FQ = 41) and prefabricated systems (FQ = 66), where robotic applications (FQ = 26) and 3D printing (FQ = 8) are increasingly bridging the gap between digital design and physical execution. Notably, the rising emphasis on “digital transformation” (FQ = 16) reflects China’s strategic alignment with global Industry 4.0 frameworks while adapting them to its unique urbanization challenges.

Download:

• PNG
  larger image
• TIFF
  original image

Table 14. Cluster statistics of high-frequency keywords.

https://doi.org/10.1371/journal.pone.0326505.t014

Parallel to technological breakthroughs, the research frontier highlights systemic industrial upgrading through “construction industrialization” (FQ = 88) and “high-quality development” (FQ = 31). Emerging priorities such as full lifecycle management (FQ = 15) and green construction (FQ = 22) underscore a dual focus on process optimization and sustainability. However, the fragmented adoption of collaborative development models (FQ = 39) and persistent gaps in cross-industry data integration (implied by low “standardization” FQ = 6) signal unresolved challenges in achieving true industry-wide digital cohesion.

Policy frameworks play a pivotal yet evolving role. While initiatives like “new construction industrialization” (FQ = 22) demonstrate effective alignment with technological goals, the underdeveloped financial incentives for SMEs (implied by FQ = 5) and mismatches in talent cultivation (FQ = 30) reveal critical gaps in policy instrument design. This disconnect is further evidenced by the PMC index analysis, which identifies inconsistencies in policy coherence despite post-2020 improvements.

Looking ahead, three interconnected trajectories will define China’s intelligent construction evolution:

Firstly, Deepening human-technology symbiosis through AI-augmented workforce training and human-robot collaboration systems, addressing the “talent training model” (FQ = 30) challenges while scaling up computer vision (FQ = 4) and IoT-enabled safety solutions (FQ = 23).

Secondly, Policy-innovation feedback loops translate research insights into adaptive governance, particularly in standardizing BIM-Digital Twin integration and creating dynamic incentive mechanisms for SME participation.

Finally, Ecosystemic sustainability via closed-loop material flows powered by robotic prefabrication and blockchain-tracked supply chains, operationalizing “green construction” (FQ = 22) concepts at scale.

The path forward demands tighter coupling between academic research, industry implementation, and policy experimentation. As China advances toward its 14th Five-Year Plan targets, success will hinge on transforming current technological leadership in areas like parametric design (FQ = 7) and smart management platforms (FQ = 5) into comprehensive, policy-supported ecosystems that redefine global construction standards.

Summary and the findings of the research

This study provides a systematic evaluation of CICP from 2010 to 2022, utilizing the PMC index model to assess the formulation quality, evolution, and effectiveness of 30 national policies. The findings reveal that:

1. (1) CICP demonstrates satisfactory overall quality, with an average PMC score of 6.93 (“Acceptable”). The policies exhibit distinct phased development characteristics, divided into the Cultivation and Exploration Phase (CEP: 2010–2019) and the Development and Promotion Phase (DPP: 2020–2022). The CEP focused on foundational digital infrastructure and technological innovation, such as BIM and IoT, but faced challenges in practical application and policy continuity. In contrast, the DPP achieved higher policy quality, with an average PMC score of 7.03 (“Positive”), and saw the emergence of “Excellent” policies like P15 and P25. These policies emphasized industrialization, digitalization, and sustainability, driving rapid advancements in intelligent construction.
2. (2) The evolution of CICP marks a transition from fragmented, exploratory strategies to integrated, goal-oriented frameworks. Key factors driving this shift include technological advancements, accumulated implementation experience, and heightened global competition. The DPP’s effectiveness stems from its ability to align policy tools with strategic goals, fostering scalable and sustainable outcomes. Lessons for future policymaking highlight the importance of adaptive governance, integrating technological trends with institutional capacity, and fostering innovation ecosystems that bridge academia, industry, and government.
3. (3) The study identifies persistent gaps in policy scope, effectiveness, and incentive mechanisms. The PMC surface analysis of high-rated policies (P15, P17, P25, and P27) reveals deficiencies in areas such as policy scope, regulatory perspectives, and implementation frameworks. These gaps underscore the need for further refinement to achieve near-optimal policy formulation quality. Additionally, the research frontier analysis indicates that China’s intelligent construction sector is undergoing a transformative phase driven by technological convergence and policy innovation, with emerging priorities such as full lifecycle management and green construction. Future trends suggest a deepening of human-technology symbiosis, policy-innovation feedback loops, and ecosystemic sustainability.

Recommendations about CICP

1. (1) Government policy optimization framework

The government should adopt a holistic approach to policy formulation, integrating top-down strategic planning with bottom-up market incentives. This includes fostering public-private partnerships (PPPs) and leveraging blockchain-enabled transparency in procurement. Adaptive governance structures, such as AI-powered policy simulators, should be established to forecast implementation challenges and optimize resource allocation.

1. (2) Addressing policy gaps

Expanding Policy Scope. Policies should transcend the current government-centric paradigm by integrating market-driven mechanisms. For instance, performance-linked tax rebates could be introduced for enterprises adopting sustainable practices, such as closed-loop material flows or AI-driven carbon accounting systems. Hybrid governance models, including outcome-based contracting, should be developed to incentivize private sector participation in intelligent construction projects.

Strengthening Policy Effectiveness. To bridge the gap between high-level objectives and on-ground execution, policies should adopt granular, metrics-driven implementation frameworks. A dynamic monitoring system should be established to quantify progress in key areas, such as BIM-Digital Twin integration and robotic prefabrication. Mandatory lifecycle carbon audits for state-funded projects and real-time data-sharing mandates across stakeholders via interoperable platforms could enhance accountability and transparency.

Broadening Policy Perspectives. Policies must address cross-sectoral challenges by dismantling silos between technological advancement and socio-industrial transformation. Cross-ministerial task forces should be formed to standardize IoT protocols for construction robotics and align vocational training with emerging skill demands, such as AI-augmented design. Additionally, global alignment with ISO standards for intelligent construction should be pursued while tailoring frameworks to China’s unique urbanization challenges.

Diversifying Incentives. Moving beyond traditional subsidies, policies should introduce multidimensional motivators that catalyze systemic change. Tiered R&D tax credits could be offered to enterprises adopting Industry 4.0 technologies, such as 3D printing and robotic automation. Green financing mechanisms, such as “intelligent construction bonds,” should be launched to attract private capital into sustainable infrastructure projects.

1. (3) Future optimization directions

Future policies should focus on deepening human-technology symbiosis through AI-augmented workforce training programs and human-robot collaboration systems. Ecosystemic sustainability should be promoted by operationalizing circular economy principles, such as blockchain-tracked supply chains and robotic prefabrication. Strengthening ties between academia and policymakers will be critical to translating frontier research (e.g., digital twins, smart materials) into actionable policies.

Implementing these recommendations would enable to address current policy gaps, align CICP with global technological frontiers, and solidify its leadership in intelligent construction, ensuring sustainable and high-quality development.

Strengths and limitations

This paper makes several unique contributions to the field of CICP research. First, it provides scientific insights into the evolution, structural characteristics, and formulation quality of CICP, offering data-based suggestions for policy optimization. Second, the research introduces an innovative dynamic and comparative approach that incorporates a temporal dimension—a key element often overlooked in traditional policy evaluations. By considering the time-sensitive nature of policy formulation and implementation, this study provides a more relevant and nuanced framework for assessing and improving policies over time. The PMC index model used in this study offers a fresh perspective on policy evaluation, enabling policymakers to track the progression and effectiveness of intelligent construction policies quantitatively. This methodological approach not only fills a gap in existing literature but also enhances the reliability and objectivity of policy assessments, setting a new standard for future research.

However, methodological diversity characterizes policy evaluation research. There exists no universally applicable criterion for selecting policy evaluation variables across different studies, as variable determination fundamentally depends on specific research contexts, theoretical frameworks, and policy implementation environments. Critically, this research focuses on national policies in China, emphasizing macro-level features with detailed exploration. While this approach offers valuable insights into the broader policy landscape, it also comes with some limitations. The central government’s strong directives have influenced regional authorities to introduce a range of policies aimed at stimulating intelligent construction initiatives in recent years. However, the analysis of national policies, though comprehensive, may overlook the practical implementation challenges at the local level. For example, regional variations in economic conditions, institutional capacities, and enforcement mechanisms can result in differing outcomes despite alignment with national directives.

Moreover, the reliance on textual data to assess policy content may fail to capture the nuanced, qualitative aspects of policy execution, such as local adaptation processes and the real-world challenges faced by stakeholders. These limitations should be considered when interpreting the results of this study. Recognizing the creation, implementation, and performance of these policies at both the national and regional levels is crucial, as local policies aligned with national directives can stimulate diverse and new research inquiries within the CICP domain.

Future research could address these gaps by incorporating qualitative methods or case studies, including:

Longitudinal Studies. Future studies could incorporate longitudinal research to explore the long-term impacts of intelligent construction policies, assessing how policies evolve and their sustained effects over time.

Qualitative Research. Further research could involve qualitative studies that collect feedback from stakeholders in the construction industry. This would provide deeper insights into the real-world challenges of policy implementation and complement the quantitative analysis presented in this paper.

Cross-Regional Comparisons. Expanding research to compare policies across different regions could reveal how local contexts influence policy outcomes, offering valuable lessons for scaling up intelligent construction initiatives nationwide.

These future research directions will help build upon the findings of this paper, providing a more comprehensive understanding of how to optimize intelligent construction policies in a rapidly evolving field.