Prioritizing Core Data Sets for Smart City Governance: Evidence from Thirty-Six Cities in Thailand

Highlights

What are the main findings?

• A large and systematic gap exists between governance demand for smart-city data and the operational availability of data across all seven smart-city domains in Thailand.
• A small Core Common Data Set (6.7% of all assessed data properties) accounts for most governance-critical data needs, clustering around population, geospatial, and planning/performance information.

What are the implications of the main findings?

• Smart-city performance depends more on strengthening data governance, metadata, and stewardship than on expanding digital platforms or sensors alone.
• The NPI–Coverage–PR framework provides a practical tool for prioritizing data investments toward the datasets that generate the greatest policy and governance value.

Abstract

Smart city initiatives increasingly rely on interoperable and high-quality urban data, yet many cities lack systematic methods for prioritizing which datasets should be developed first. This study proposes an evidence-based framework for smart city data prioritization that integrates data need, data availability, and policy urgency into a unified decision-support model. Using standardized data elements across seven nationally defined smart city domains, the framework was applied to thirty-six certified smart cities in Thailand. Data were collected from municipal authorities and national platforms and structured using ISO-based data element and metadata principles. For each data element, a Need Priority Index, Coverage score, and Policy Readiness indicator were computed to assess governance-relevant data readiness. The results reveal a persistent imbalance between high data demand and low data availability across all domains, with Smart Mobility, Smart Living, Smart Energy, and Smart Economy showing the highest urgency. A Core Common Data Set representing 6.7% of assessed properties was identified, centered on population data, geospatial infrastructure, and plans and performance indicators. The framework provides a scalable approach for guiding investments in interoperable smart city data systems.

1. Introduction

Across the global smart city landscape, data functions not merely as a technical resource but as a form of urban infrastructure that enables cities to plan, operate, and evaluate public services [1]. As urbanization accelerates, with the United Nations projecting that 68 percent of the world’s population will reside in urban areas by 2050, cities face growing pressure to manage environmental quality, mobility systems, public safety, and social welfare through increasingly data-driven governance [2]. This pressure is reflected in the rapid proliferation of smart city initiatives worldwide, illustrating how governments are embedding data into the core of urban management and service delivery [3,4,5]. Recent systematic reviews further confirm that urban digitalization is increasingly driven by data platforms, analytics, and information infrastructures, rather than by isolated applications alone [4,5].

High-quality, well-defined, and governable data enables local authorities to monitor urban conditions, coordinate services, and design responsive policy interventions. Yet in many developing and emerging economies, smart city ambitions continue to outpace data readiness. Fragmented data systems, inconsistent definitions, weak metadata practices, and limited institutional interoperability constrain the ability of cities to translate digital investments into effective governance outcomes [6,7,8,9]. Thailand exemplifies these challenges. Although 36 cities across 25 provinces have been certified as smart cities by the Digital Economy Promotion Agency (DEPA), their data ecosystems remain uneven, with substantial variation in the existence, coverage, and operational usability of key datasets across smart city domains [10].

Importantly, this unevenness does not stem simply from data absence. Smart city data in Thailand is distributed across multiple national and international frameworks, including global indicator systems, ASEAN planning guides, and national city data platforms, which were developed for different purposes and therefore lack common semantic and metadata structures [11,12,13,14]. Similar fragmentation has been documented internationally, where heterogeneous indicator systems and sector-specific platforms prevent data from being operationalized for integrated urban governance [11,15,16]. International standards such as ISO/IEC 11179 (metadata registries), ISO 19115 (geospatial metadata), and ISO 8000 (data quality) emphasize that data must be defined in terms of standardized data elements, properties, and value domains in order to be interoperable across systems [17,18,19]. Without such harmonization, smart city data ecosystems remain structurally fragmented and unable to support cross-domain governance and decision-making [20,21].

Existing smart city frameworks and national data platform initiatives provide extensive domain-level and indicator-level guidance, but they do not offer a quantitative and operational method for prioritizing data elements based on actual city-level readiness [6,11,12,13,22,23]. Most frameworks specify what data should exist, but not which data should be prioritized first when resources, institutional capacity, and data availability are constrained. International governance models similarly emphasize interoperability and stewardship yet rarely provide empirical mechanisms for sequencing data investments under real-world conditions [9,15,24]. As smart city functions increasingly depend on cross-domain information flows linking mobility, environment, governance, economy, and social services [1,11], the absence of an evidence-based prioritization mechanism limits cities’ ability to design data platforms, allocate resources, and establish governance structures aligned with operational realities.

These challenges are further amplified by geographic, economic, and institutional diversity. Data needs to differ substantially across cities exposed to distinct environmental risks, demographic pressures, and infrastructural conditions. Comparative smart city research shows that regional and metropolitan contexts strongly shape both data requirements and governance capacity [6,25,26]. Consequently, a uniform national checklist of datasets cannot adequately reflect contextual variation, nor can it resolve the structural fragmentation of data sources across agencies and platforms. What is required is a context-sensitive, quantitative method that distinguishes understanding what data is needed, what data exists, and where the largest readiness gaps occur.

The NPI–Coverage–PR framework provides such a mechanism. Building on data governance and metadata-based interoperability theory [16,17,20,27], this study operationalizes smart city data readiness through three interrelated dimensions: (1) Need Priority Index (NPI), measuring the governance importance of standardized data properties; (2) Coverage, measuring whether data exists in official administrative or digital systems at the city level; and (3) Policy Readiness (PR), identifying high-need, low-coverage conditions that require urgent intervention. By applying this framework to 1224 standardized data properties grouped under 124 data elements across seven nationally defined smart city domains—Smart Environment, Smart Mobility, Smart Living, Smart People, Smart Energy, Smart Economy, and Smart Governance—this study moves beyond abstract indicator lists to provide a reproducible, governance-oriented method for prioritizing urban data investments.

Accordingly, this study addresses two interrelated evidence gaps. First, it quantifies the structural imbalance between data demand and data availability across Thailand’s certified smart cities, identifying where critical datasets are missing from operational systems. Second, it examines cross-domain and spatial variation in data readiness, providing an empirical basis for context-sensitive data governance. By linking standardized data properties to city-level readiness through the NPI–Coverage–PR framework, the study derives a Core Common Data Set (CCDS) that identifies the minimum data backbone required to support interoperable, policy-relevant smart city governance across heterogeneous urban contexts.

2. Related Work and Framework Positioning

Smart city development is increasingly recognized as a data-intensive and governance-driven process in which urban services, infrastructure, and policy systems depend on interoperable, well-governed data [1,6]. International research shows that smart cities succeed not through isolated technological deployments, but through the construction of integrated data ecosystems that enable cross-sector coordination, spatial analytics, and performance monitoring [28,29,30]. Recent systematic reviews further confirm that data architectures, interoperability frameworks, and governance mechanisms are now central to smart city implementation [4,5].

2.1. Global and Regional Smart City Data Frameworks

Numerous international initiatives have sought to formalize the indicators and datasets required to support smart city governance. Sharifi’s synthesis of more than thirty global smart city frameworks demonstrates that urban performance assessment relies on a diverse and often overlapping set of indicators across domains such as environment, mobility, economy, governance, and quality of life [6]. Similarly, the United for Smart Sustainable Cities (U4SSC) framework and the OECD Digital Government Index emphasize data interoperability, institutional readiness, and governance capacity as key enablers of digital urban transformation [11,12].

At the regional level, the ASEAN Smart Cities Framework and the ASEAN Smart City Planning Guidebook provide strategic guidance for aligning urban development with regional priorities [13,14]. However, these frameworks focus primarily on policy domains and development objectives rather than on how standardized data elements should be defined, harmonized, and prioritized across heterogeneous cities. Comparative research confirms that such high-level frameworks often lack operational mechanisms for resolving semantic inconsistencies and data fragmentation across institutions [15,16].

2.2. National Smart City Data Initiatives in Thailand

Thailand has established an extensive set of smart city policies, digital platforms, and technical guidelines through the Digital Economy Promotion Agency, including the City Data Platform Development Framework and national dataset catalogs [22,23]. These initiatives define datasets, metadata fields, and platform architectures intended to support municipal data sharing and analytics. However, as in many other countries, these platforms primarily function as data repositories rather than as fully governed information infrastructures. International digital government research shows that cataloging datasets is insufficient unless accompanied by standardized data elements, shared semantics, and formal stewardship mechanisms [9,24]. As a result, cities may technically “have data” while remaining unable to integrate, compare, or reuse it for cross-domain governance.

2.3. Data Governance, Metadata, and Interoperability

From a data-governance perspective, effective smart city systems require not only data availability but also semantic interoperability, metadata harmonization, and clearly defined data stewardship roles. International standards such as ISO/IEC 11179, ISO 19115, and ISO 8000 provide formal structures for defining data elements, value domains, and metadata attributes so that data can be consistently interpreted and exchanged across systems [17,18,19]. Without these foundations, urban data platforms risk becoming collections of incompatible datasets rather than integrated governance infrastructures [20].

Recent research in smart cities and information systems emphasizes the importance of ontology-based and semantic frameworks for integrating heterogeneous urban data sources [20,21]. These approaches demonstrate how GIS, IoT, administrative, and planning data can be unified through shared data models, enabling cross-domain analytics and real-time applications. However, most of this work focuses on technical integration rather than on governance-oriented prioritization of which data elements should be developed first.

2.4. Limitations of Existing Prioritization Approaches

Although global and national frameworks provide extensive lists of indicators and datasets, few offer quantitative methods for prioritizing data based on real-world city-level readiness. Smart city indices and benchmarking systems focus on what should be measured, but rarely assess whether cities possess the institutional, technical, and metadata capacity to produce and govern those data [6,11]. Similarly, governance-oriented frameworks propose rules and principles for data management but lack empirical mechanisms for ranking priorities under resource constraints [15,16].

Osu and Navarra’s data governance framework [27] advances this field by emphasizing the importance of identifying essential data elements and their properties. However, it does not provide a quantitative method for comparing readiness across cities or for linking data need to operational availability. To clarify how the present study extends existing work,

Table 1. Positioning of the NPI–Coverage–PR framework relative to existing smart city data and governance frameworks.

Framework/Source	Primary Focus	Level of Analysis	How Data Is Conceptualized	What It Provides	What It Does Not Provide
Data management and governance for smart cities [27].	Principles of data governance, quality, and integration	City/ Agency	Data as managed assets requiring quality, security, and compliance	Conceptual guidance on data governance, attributes, and management processes	No quantitative prioritization; no measurement of data need, availability, or urgency
Global synthesis of smart city indicators [11].	Comparative assessment and taxonomy of smart city indicators	Cross-city	Data as indicators for benchmarking and evaluation	Comprehensive mapping of global smart city indicator sets	Descriptive only; no operational prioritization; no linkage between need and readiness
ASEAN Smart Cities Framework [12] and MLIT Planning Guidebook [13].	Strategic and policy guidance for smart city development	City/ Region	Data as supporting evidence for planning and coordination	Policy orientation, regional alignment, development pathways	No specification of data elements, metadata, or readiness; no prioritization logic
Thailand City Data Platform (CDP) guidelines [14,22,23].	Technical architecture and national data catalogs	City/ Agency	Data as datasets and system components	Dataset lists, platform design, and system-level guidance	Does not connect data supply to data need; no Gap or urgency analysis
This study (NPI–Coverage–PR)	Evidence-based prioritization of smart city data	36 Thai smart cities across regions	Data as standardized, policy-relevant elements with measurable readiness	Quantifies data need, availability, gaps, and urgency across domains using ISO-aligned data structures	Focused on certified Thai smart cities; detailed metadata profiles (e.g., DCAT-AP, DCAT-3) reserved for future work

Note: Existing frameworks conceptualize smart city data either as policy guidance, technical inventories, or benchmarking indicators. None integrate data need, data availability, and policy urgency into a single operational model. The NPI–Coverage–PR framework fills this gap by linking standardized data elements with readiness metrics to support evidence-based data governance and investment decisions.

positions the NPI–Coverage–PR framework relative to major international, regional, and national smart city data and governance frameworks.

2.5. Positioning the NPI–Coverage–PR Framework

The NPI–Coverage–PR framework addresses these limitations by integrating ISO-based data element modeling with a readiness-oriented prioritization logic. By quantifying perceived data importance (NPI), operational availability (Coverage), and policy urgency (PR), the framework transforms descriptive indicator systems into an operational decision-support tool. It enables policymakers to distinguish between data that is theoretically required, data that is operationally available, and data that is urgently needed for governance. In doing so, the framework bridges three previously disconnected strands of research: global smart city indicator systems [6,11], national data platform initiatives [22,23], and data governance and interoperability theory [16,17,18,19,20,21]. The result is an empirically grounded method for identifying a Core Common Data Set capable of supporting interoperable, scalable, and policy-relevant smart city data.

3. Methods

3.1. Research Design

This study employed a quantitative, cross-sectional design to evaluate data need, data availability, and policy readiness across Thailand’s 36 certified smart cities. The methodological approach integrates descriptive statistics, nonparametric group comparisons, and correlation analysis within a data governance and metadata-based interoperability framework [27]. Smart city data readiness is assessed at the level of standardized data elements and their properties, rather than at the level of unstructured datasets or ad hoc indicators, in accordance with ISO/IEC 11179 (metadata registries) and ISO 19115 (geospatial metadata). This design enables systematic comparison of heterogeneous cities while preserving semantic and structural consistency across data domains.

3.2. Study Sample and Data Sources

The study sample comprises 36 officially certified smart cities in 25 provinces, covering all seven national smart city domains: Smart Environment, Smart Mobility, Smart Living, Smart People, Smart Governance, Smart Energy, and Smart Economy. Certification by the DEPA ensures that these cities represent Thailand’s institutional smart city frontier, where formal data governance and digital infrastructure are actively being implemented.

Data was obtained from four complementary sources: (1) municipal administrative records; (2) provincial digital development units; (3) national and city-level digital platforms and dashboards; and (4) a structured questionnaire completed by designated smart city coordinators or local digital-government officers.

The questionnaire (Supplementary File S1) operationalizes data status, data properties, and governance need for 1224 data properties across 124 standardized data elements. Using multiple sources reduces institutional bias and enables triangulation between administrative records, technical systems, and governance assessments.

3.3. Data Model and Database Structure

All smart city data were structured using international metadata and data element principles. Data were organized hierarchically into domains, standardized data elements, and properties, consistent with ISO/IEC 11179 and ISO 19115 definitions of data elements, value domains, and metadata attributes.

Two complementary datasets were constructed. First, a property-level dataset consisting of 44,064 records (36 cities × 1224 properties) was used to compute Need Priority Index (NPI), Coverage, Gap, and Policy Readiness (PR) for each standardized data property. Second, a city-level dataset aggregated domain-wise indicators and incorporated provincial and regional identifiers for spatial and comparative analysis. This dual structure enables fine-grained readiness assessment while preserving compatibility with international data governance standards.

3.4. Applicable-Only Approach

Cities differ in their institutional mandates and service portfolios; therefore, not all data elements are relevant to all municipalities (e.g., some cities do not operate public transport systems or manage industrial energy facilities). To avoid structural bias, the study applies an applicable-only approach. For each data element and property, cities were included in calculations only when the item was logically relevant and a valid response was provided.

Triangulation was applied when discrepancies arose between platform data, administrative records, and expert reports. In such cases, a precedence rule was used (platform-verified data > administrative databases > expert reports), followed by follow-up verification with city coordinators when needed. This ensures that Coverage reflects operationally defensible data existence rather than subjective perception.

3.5. Measures and Indicator Construction

(1)

Need Priority Index (NPI). Data need was measured using a three-level policy-priority scale in the questionnaire (Supplementary File S1): High need = 3, Low need = 2, No need = 1. Scores were averaged across applicable cities and normalized to a 0–1 scale to produce the NPI.

(2)

Coverage. Coverage measures whether a data property exists in official administrative or digital systems at the city level. Respondents reported whether each property was available (1), not available (0), or not applicable. Coverage was calculated as the proportion of applicable cities reporting availability. Coverage therefore reflects operational digital existence, not platform interoperability or data quality.

(3)

Gap. The Gap index measures the structural mismatch between governance demand and data availability: Gap = NPI − Coverage.

(4)

Policy Readiness (PR). PR captures urgency for data development: PR = NPI × (1 − Coverage). High PR values indicate high-need, low-availability conditions that constrain effective governance.

3.6. Statistical Procedures

All indicators were computed at the property level using the applicable-only approach. Four analytical components were implemented:

(1)

Descriptive assessment of data readiness. Descriptive statistics were used to summarize NPI, Coverage, Gap, and PR across the seven smart city data domains. These indicators quantify, respectively, perceived governance need, operational data availability, structural data deficits, and urgency for policy intervention. Domain-level means and distributions provide a national profile of Thailand’s smart city data readiness, enabling identification of domains in which data infrastructure and metadata governance are most critically underdeveloped.

(2)

Regional comparison of data needs. To examine whether data needs reflect contextual differences across Thailand, Kruskal–Wallis nonparametric tests were conducted to compare NPI scores across five geographic regions (North, Northeast, Central, East, and South). This test was selected because NPI values are ordinal derived and do not assume normality. Effect sizes were reported using epsilon-squared (ε²), which quantifies the proportion of variance in data needs attributable to regional context. This analysis supports data governance planning by identifying regions where structural, environmental, and socio-economic conditions generate systematically higher data requirements.

(3)

Cross-domain co-occurrence of data needs. Spearman’s rank-order correlation coefficients (ρ) were used to assess the co-occurrence of data needs across smart city data domains. This analysis reveals whether data requirements in one domain (e.g., Smart Living) are systematically associated with those in others (e.g., Smart Environment or Smart Governance), thereby indicating interdependence that must be considered when designing interoperable data architectures and cross-sectoral governance mechanisms. All correlations were computed using pairwise applicable-only cases to avoid distortion from structurally non-applicable data elements.

(4)

PR-based prioritization and governance tiering. PR distributions were used to rank smart city data domains, regions, and data elements according to urgency for intervention. High-PR values indicate conditions in which essential data are missing despite strong governance demand, signaling priority areas for investment in data infrastructure, metadata harmonization, and stewardship arrangements. PR-based tiering thus translates statistical results into actionable guidance for policymakers, enabling evidence-based allocation of resources toward the most critical gaps in the national smart city data ecosystem.

All analyses were conducted using R and Python version 2.6, ensuring full reproducibility of indicator construction, statistical testing, and prioritization outputs.

4. Findings

4.1. National Patterns of Data Need and Readiness

Across Thailand’s 36 certified smart cities, the results reveal a systematic imbalance between governance demand for data and the operational availability of data in administrative and digital systems. Using the applicable-only approach, NPI values are uniformly high across all seven smart city domains, ranging from 0.769 to 0.890, indicating strong and consistent agreement among local smart-city administrators that these data are essential for governance. In contrast, Coverage remains critically low, with domain-level means ranging from 0.110 to 0.231 (

Table 2. Average scores across smart city development dimensions.

Smart City Data Domains	Coverage (Mean)	NEED/PGI (Mean)	Gap (=PGI − Coverage)	PR (=PGI × (1 − Coverage))
Smart environment	0.231	0.769	0.537	0.591
Smart energy	0.154	0.846	0.691	0.715
Smart economy	0.162	0.838	0.676	0.702
Smart governance	0.184	0.816	0.632	0.666
Smart mobility	0.110	0.890	0.781	0.793
Smart living	0.150	0.850	0.699	0.722
Smart people	0.163	0.837	0.674	0.701

Note: High PGI values reflect a ceiling effect, as many respondents assigned the maximum importance score. Therefore, Gap and PR are used as the primary prioritization metrics, as they identify data properties with high governance demand but low operational availability.

), indicating that many required data properties do not yet exist in usable digital or administrative form at the city level.

This divergence produces large structural readiness gaps. Gap values range from 0.537 to 0.781, while PR values range from 0.591 to 0.793, confirming that high governance demand coincides with weak operational data availability. Using PR as an integrated urgency metric (PR ≥ 0.70 = critical; 0.50–0.69 = high; <0.50 = moderate), four domains fall in the critical tier: Smart Mobility (PR = 0.793), Smart Living (0.722), Smart Energy (0.715), and Smart Economy (0.702). Smart Governance (0.666), Smart People (0.701), and Smart Environment (0.591) fall within the high-urgency tier.

PR values exhibit moderate dispersion across cities (interquartile ranges approximately 0.10–0.18 across domains), indicating that high urgency is not driven by a few outliers but reflects a national-level structural condition. These results show that Thailand’s smart-city data challenges are systemic rather than idiosyncratic, and that data readiness constraints affect nearly all domains of urban governance.

4.2. Regional Variation in Data Needs

Kruskal–Wallis tests indicate statistically significant regional differences in data needs (PGI/NPI) across all seven smart city domains (p < 0.05), with medium-to-large effect sizes (ε² = 0.313–0.651;

Table 3. Comparison across five geographic regions using the Kruskal–Wallis test.

Indicator	χ2	df	p	ε2
PGI_Smart environment	16.5	4	0.002	0.471
PGI_Smart energy	11.9	4	0.018	0.340
PGI_Smart economy	14.0	4	0.007	0.399
PGI_Smart governance	16.5	4	0.002	0.471
PGI_Smart mobility	11.0	4	0.027	0.313
PGI_Smart living	22.8	4	<0.001	0.651
PGI_Smart people	12.8	4	0.012	0.366

Note: χ² = Kruskal–Wallis statistic; df = degrees of freedom; p = significance level; ε² = effect size. Effect-size benchmarks: small ≈ 0.01–0.08, medium ≈ 0.09–0.24, large ≥ 0.25. Results computed using applicable-only cases.

). These values imply that 31–65% of the variance in perceived data need is associated with regional context rather than random variation.

The strongest regional differentiation occurs in Smart Living (ε² = 0.651), Smart Environment (ε² = 0.471), and Smart Governance (ε² = 0.471), reflecting differences in demographic structure, environmental exposure, and administrative complexity across Thailand. Northern and northeastern regions exhibit higher median demand for environmental and health-related data, whereas eastern and southern regions show elevated needs for mobility, energy, and economic data, consistent with their industrial, logistics, and tourism profiles.

These findings demonstrate that smart-city data needs are context-dependent, reinforcing the need for region-sensitive data governance rather than uniform national dataset templates.

4.3. Provincial-Level Comparison

At the provincial level, descriptive differences in PGI are visible across the 25 provinces, but none of the Kruskal–Wallis tests reach statistical significance (p = 0.102–0.280;

Table 4. Provincial level comparison of data needs (PGI) across 25 provinces using the Kruskal–Wallis test.

Indicator	χ2	df	p	ε2
PGI_Smart environment	32.4	24	0.117	0.926
PGI_Smart energy	29.3	24	0.209	0.837
PGI_Smart economy	29.3	24	0.211	0.836
PGI_Smart governance	31.6	24	0.138	0.902
PGI_Smart mobility	27.5	24	0.280	0.787
PGI_Smart living	33.1	24	0.102	0.946
PGI_Smart people	27.7	24	0.274	0.791

Note: χ² = Kruskal–Wallis statistic; df = degrees of freedom; p = significance level; ε² = effect size. Lack of statistical significance (p > 0.05) results from small group sizes and many comparison groups, despite large effect-size estimates.

). This reflects two structural features of the dataset. First, most provinces contain only one certified smart city, limiting within-group variance and statistical power. Second, the large number of comparison groups inflates the critical χ² threshold, making statistically significant differences unlikely even when dispersion is substantial.

Although ε² values appear numerically large, these should not be interpreted as reliable estimates of provincial effects under such sparse group conditions. Provincial-level results are therefore descriptive rather than inferential and should not be used for policy targeting. Regional-level aggregation provides a more stable and interpretable scale for smart-city data governance and investment planning.

4.4. Cross-Domain Co-Occurrence of Data Needs

Spearman’s rank correlations reveal strong and systematic co-occurrence patterns in data needs across smart city data domains (

Table 5. Spearman’s rank-order correlations (ρ) of PGI across the seven smart city data domains (pairwise applicable-only).

ρ (PGI)	Smart Living	Smart People	Smart Environment	Smart Governance	Smart Energy	Smart Economy	Smart Mobility
Smart living	—	0.615 *	0.827 ***	0.718 ***	—	—	0.309 †
Smart people		—	0.619 ***	0.595 ***	—	—	—
Smart environment			—	0.538 *	—	0.254 (ns)	0.460 *
Smart governance				—	0.452 *	0.326 †	0.379 *
Smart energy					—	0.515 **	—
Smart economy						—	—
Smart mobility							—

Note: ρ = Spearman’s rank correlation. Pairwise applicable-only calculation used to avoid bias from structural NA values. * p < 0.05, ** p < 0.01, *** p < 0.001, † marginal significance (p < 0.10), ns = not significant. “—“ indicates insufficient non-NA pairs to compute correlation.

). The strongest cluster links Smart Living, Smart Environment, Smart Governance, and Smart People (ρ = 0.595–0.827, p < 0.001), indicating that cities requiring data to support quality of life and social services also demand environmental and governance information. This cluster reflects the interdependence of human well-being, ecological conditions, and administrative capacity in urban systems.

A second cluster links Smart Energy and Smart Economy (ρ = 0.515, p = 0.001), highlighting the coupling between energy consumption, industrial activity, and economic performance. Smart Mobility also correlates with Smart Environment (ρ = 0.460, p = 0.005), reflecting the environmental impacts of transport systems. These patterns imply that data should be designed and governed as interoperable bundles rather than as isolated datasets, since policy-relevant information flows span multiple smart city data domains.

4.5. PR-Based Prioritization of Data Gaps

PR values provide a governance-oriented ranking of urgency by integrating high data need with low availability. Across all domains, most PR scores exceed 0.60, indicating widespread high to critical urgency. Using PR thresholds, Smart Mobility, Smart Living, Smart Energy, and Smart Economy fall into the critical tier, while Smart Governance, Smart People, and Smart Environment fall into the high tier.

At the regional level, cities in the Northeast and South exhibit systematically higher PR values across multiple domains, indicating that these regions face the greatest structural constraints in data readiness. These results translate statistical patterns into actionable guidance, enabling policymakers to prioritize investment in metadata harmonization, geospatial infrastructure, and real-time data systems where governance needs are most acute (

Table 6. Prioritization of properties of data elements across the seven smart city data domains (pr-based classification).

Core Layer	Examples of High Priority Properties of Data Element Across All Seven Smart City Data Domains	Policy Rationale/ Governance Logic
Layer 1: structural core	- Housing types - Land-use area by type - Business survival rate - SME share of GDP - ICT equipment data - Types of electricity users	- Provides baseline for city and sectoral planning - Supports strategic planning for economy, ICT, land use, and energy - Establishes minimum structural datasets required for interoperability - Ensures consistent reference layers across smart city data domains.
Layer 2: operational core	- Real-time traffic information - Traffic volume - Real-time air - Quality reporting; update frequency - Load and storage data - Smart-meter indicators - Online petitions - Digital participation - Adult learning registration - Mobile e-learning usage	- Enables real-time and near-real-time city management - Critical for agile operations in mobility, environment, energy, and civic engagement; supports rapid response to disruptions - Strengthens responsiveness of digital public services.
Layer 3: functional core	- Number of jobs by sector - Tourist numbers - Tourism plans - Health facilities and their locations - Learning resources - Smart City Plan - Public health plans - Implementation plans	- Supports area and sector specific policy and service design - Enables targeted interventions and context-sensitive planning - Guides to resource allocation based on population and spatial patterns - Reinforces long-term urban development strategies.

).

4.6. Identification of the Core Common Data Set (CCDS)

To identify a national Core Common Data Set, properties of data elements were classified as core if they simultaneously exhibited high policy urgency (PR) and limited operational availability (Coverage < 0.60). This threshold reflects a conservative criterion for identifying datasets that are both critically needed and structurally missing. Sensitivity checks using alternative cut-offs (0.50 and 0.70) produced similar core sets, indicating robustness of the classification.

Out of 1224 properties of data elements, 82 (6.7%) meet the core criteria, forming a compact but strategically important CCDS. These core items cluster into three foundational pillars: population data, geospatial infrastructure, and plans and performance indicators. Together, these pillars provide the minimum interoperable data backbone required for cross-domain governance, spatial integration, and policy monitoring. The fact that fewer than 7% of assessed properties qualify as core highlights both the feasibility and the urgency of focusing national data investments on a small, high-impact subset of standardized datasets. (

Table 7. Number of smart city foundational data items classified as core and extended across the seven domains.

Smart City Data Domains	Core	Extended	Total	Core_%	Extended_%
Smart environment	22	160	182	12.09	87.91
Smart mobility	12	148	160	7.50	92.50
Smart governance	13	171	184	7.07	92.93
Smart people	11	148	159	6.92	93.08
Smart energy	7	94	101	6.93	93.07
Smart economy	10	201	211	4.74	95.26
Smart living	7	220	227	3.08	96.92
Total	82	1142	1224	6.70	93.30

).

4.7. City-Level Illustration: Khon Kaen

To illustrate how the NPI–Coverage–PR framework translates into operational data-governance challenges at the city level, Khon Kaen is presented as an example. As one of Thailand’s major regional cities and an early adopter of smart city initiatives, Khon Kaen exhibits strong demand for data across multiple smart city domains but also substantial gaps in interoperable data availability.

In Khon Kaen, the highest PR values are observed in Smart Mobility, Smart Living, and Smart Governance. Within Smart Mobility, high PR is driven by strong demand for real-time traffic flow, public transport utilization, and congestion data, combined with low coverage of standardized and geocoded datasets capable of supporting integrated transport management. Although various transport-related data exists across municipal agencies and national platforms, the lack of harmonized spatial reference layers and update cycles prevents these data from being used effectively for operational planning and public service delivery.

In the Smart Living domain, Khon Kaen shows particularly high PR for health service locations, emergency response times, and population vulnerability indicators. These data are critical for supporting aging populations, disaster response, and equitable access to services. However, low coverage reflects the absence of interoperable datasets linking demographic information, facility locations, and service performance indicators, limiting the city’s ability to deploy data-driven social and public health interventions.

Smart Governance also exhibits high PR in Khon Kaen, particularly for digital participation, service request tracking, and performance monitoring indicators. While some administrative data are collected internally, they are not consistently standardized, shared, or integrated across departments, reducing their value for transparency, accountability, and citizen-centered governance.

Taken together, Khon Kaen’s profile demonstrates that high data need does not automatically translate into data readiness. The city’s experience reflects a broader national pattern in which fragmented data sources, weak metadata practices, and limited interoperability constrain the operational use of data for smart city governance. This vignette underscores the practical relevance of the CCDS and PR-based prioritization approach, showing how national-level metrics map directly onto concrete governance challenges faced by individual cities. (

Table 8. Illustrative top 10 Policy Readiness (PR) data gaps for Khon Kaen.

Rank	Smart City Data Domain	High-Priority Data Property	Governance Relevance	Typical Coverage Status
1	Smart Mobility	Real-time traffic flow	Congestion management, emergency routing, and public transport optimization	Low
2	Smart Mobility	Public transport utilization	Service planning, fare policy, and route optimization	Low
3	Smart Environment	Real-time air quality (AQI)	Public health, pollution alerts, and environmental regulation	Low
4	Smart Living	Health facility locations and capacity	Emergency response, healthcare access, and service equity	Low
5	Smart Living	Population vulnerability indicators	Disaster response, aging, and social protection	Low
6	Smart Governance	Digital service requests and complaints	Service quality monitoring and accountability	Low
7	Smart Governance	Online civic participation data	Transparency, engagement, and policy feedback	Low
8	Smart Energy	Electricity demand by user type	Load management, energy planning, and sustainability	Low
9	Smart Economy	Employment by sector	Workforce development and economic resilience	Low
10	Smart Environment	Waste generation by area	Environmental management and infrastructure planning	Low

).

5. Discussion

5.1. Structural Imbalance Between Data Need and Data Readiness

The most striking result of this study is the persistent and systemic imbalance between governance demand for data and operational data readiness across all seven smart city domains. While NPI values are consistently high, Coverage remains critically low, producing elevated PR scores nationwide. This pattern indicates that strong policy demand for data is not matched by the existence of usable data in administrative and digital systems. Similar gaps between data ambition and operational readiness have been widely reported in international smart city research, which shows that cities often invest in digital platforms without establishing the data governance, metadata, and stewardship foundations required for sustained use [1,4,8,9,31].

Low Coverage should therefore not be interpreted as a simple absence of data, but rather as a reflection of fragmented ownership, incompatible definitions, and weak metadata practices that prevent data from functioning as a shared governance resource [9,20,32]. This condition corresponds to what urban data scholars describe as structural data governance failure, in which information exists across organizations but cannot be reliably combined, compared, or reused for public decision-making [15,16]. Similar fragmentation has been documented in smart city interoperability and digital government studies [20,21].

5.2. Regional Variation and Context-Sensitive Data Governance

The significant regional differences in data needs identified in this study are consistent with comparative research showing that smart city priorities are shaped by local socio-economic, environmental, and institutional conditions rather than uniform national templates [6,25,26]. Regions exposed to greater environmental risk, demographic change, or industrial development display systematically higher demand for environmental, health, mobility, and governance data, highlighting the need for adaptive and context-sensitive data governance strategies.

These patterns also reflect the multi-level governance structure of smart cities, in which regional institutions often play a stronger role than individual municipalities in coordinating infrastructure, standards, and investment [24,33]. Digital government research similarly shows that regional and metropolitan data platforms are more effective for standardization and shared services than either fragmented local systems or rigid national mandates [9,24].

5.3. Interdependent Data Needs and the Logic of Cross-Domain Integration

The strong co-occurrence of data needs across Smart Living, Smart Environment, Smart Governance, and Smart People highlights the interdependence of social, environmental, and administrative systems in urban governance. Prior research demonstrates that policy-relevant analytics depend on integrated data across domains, not on isolated sectoral datasets [28,29,34]. Similarly, the linkage between Smart Energy and Smart Economy reflects well-established connections between energy use, industrial activity, and economic performance [30,35].

Recent metadata and semantic integration research confirm that such cross-domain dependencies require standardized data elements, shared reference layers, and metadata registries to function operationally [20,21]. The patterns observed in this study therefore reinforce the role of data models and governance structures, rather than platforms alone, as the backbone of smart city information systems [9,20].

5.4. The Core Common Data Set (CCDS) as a Governance Instrument

The identification of a CCDS comprising only 6.7% of assessed data properties aligns with international evidence that a small set of well-defined and well-governed datasets can support a wide range of urban functions when properly managed [6,15,32,36]. Population data, geospatial reference layers, and planning and performance indicators are consistently identified as foundational data assets because they enable cross-domain integration, spatial analysis, and policy accountability [6,7,28].

Urban data governance models further emphasize that prioritizing a limited set of core datasets improves scalability, reuse, and institutional coordination [16]. By deriving the CCDS empirically through PR-based prioritization, this study provides a data-driven basis for sequencing national and municipal investments toward the datasets that generate the greatest governance value.

5.5. Policy Readiness (PR) as a Decision-Support Tool

The PR index provides a practical mechanism for translating analytical results into actionable data governance priorities. Similar readiness-based prioritization approaches have been proposed in the digital government and data governance literature to guide investment in data infrastructure, metadata harmonization, and stewardship capacity [9,24,37]. The Khon Kaen illustration shows how high-PR gaps cluster around real-time mobility data, health and vulnerability indicators, and digital governance metrics—precisely the types of data that international research identifies as critical for urban resilience, equity, and integrated service delivery [1,8,38].

5.6. Broader Implications for Smart City Data Governance

Overall, the findings reinforce the view that smart city success depends less on technology deployment than on the construction of a coherent data governance architecture [8,9,25]. Investments in sensors, platforms, and analytics will remain underutilized unless supported by standardized data elements, interoperable metadata, and clearly defined stewardship arrangements [9,15,16]. By integrating ISO-based data governance principles with a readiness-oriented prioritization framework, this study provides an empirically grounded pathway for moving from fragmented data ecosystems toward policy-relevant and interoperable smart city information infrastructures. Although this analysis focuses on Thailand’s certified smart cities, the underlying governance challenges are common across many developing and middle-income countries pursuing digital urban transformation [5,6,26].

6. Conclusions

This study developed and applied the NPI–Coverage–PR framework to assess smart-city data readiness across 36 certified Thai smart cities, focusing on the alignment between data need, operational data availability, and policy urgency. By operationalizing data readiness at the level of standardized data elements and their properties, the study moves beyond descriptive indicator lists to provide an evidence-based, governance-oriented approach to smart-city data prioritization. The findings reveal a persistent and structural imbalance between data demand and operational readiness. Across all seven domains, NPI values are uniformly high, indicating strong consensus among local administrators regarding the importance of data for urban governance, while Coverage remains critically low, producing high PR values nationwide. Smart Mobility, Smart Living, Smart Energy, and Smart Economy exhibit critical levels of urgency, with Smart Governance, Smart People, and Smart Environment also displaying substantial readiness gaps. Regional analysis further shows that data needs are shaped by local socio-environmental and economic conditions, reinforcing the need for context-sensitive data governance rather than uniform national dataset templates.

A major contribution of this study is the empirical identification of a Core Common Data Set (CCDS) comprising 6.7% of assessed data properties. These core items cluster into three foundational data pillars: population data, geospatial reference data, and planning and performance indicators. Together, they represent the minimum data backbone required to support cross-domain integration, spatial analysis, and evidence-based urban governance. By deriving the CCDS through the PR metric, the study provides a replicable and transparent method for directing limited public resources toward the data assets that generate the greatest governance value.

From a theoretical perspective, this study advances smart-city research by integrating ISO-based data governance and metadata principles with a readiness-oriented prioritization framework. It demonstrates that the success of smart-city initiatives depends not merely on data availability but on the existence of standardized, governable, and institutionally embedded data elements that can function as shared information infrastructure. For policymakers and practitioners, the NPI–Coverage–PR framework offers a practical decision-support tool for sequencing investments in data platforms, metadata harmonization, and stewardship capacity according to real governance needs.

Several limitations should be acknowledged. First, data availability and importance were assessed through administrative records, digital platforms, and expert reporting, which capture operational data existence but may not reflect the full technical quality, completeness, or interoperability of underlying databases. Second, although the study covers all certified Thai smart cities, the small number of cities per province limits the statistical power of provincial-level comparisons. Third, the analysis focuses on data readiness and prioritization rather than directly measuring data quality, update frequency, or real-world use in policy processes.

Future research could extend this framework in several directions. Longitudinal studies could track how NPI, Coverage, and PR evolve as cities invest in data platforms and governance reforms. Technical audits of metadata completeness, geospatial accuracy, and interoperability could complement survey-based readiness assessments. Comparative applications in other countries or regions would further test the framework’s generalizability and inform the development of international smart-city data governance standards. Overall, this study provides a robust, transparent, and policy-relevant foundation for transforming fragmented urban datasets into a prioritized and governable smart-city data ecosystem.