A Bibliometric Analysis of Research on the Metaverse for Smart Cities: The Dimensions of Technology, People, and Institutions

Abstract

The “Metaverse” is evaluated as having significant potential in a “Smart city” design and operation. Despite growing interest, there is still a lack of comprehensive quantitative analysis on the “Metaverse”, particularly in the context of smart cities. This study conducts a bibliometric analysis of 604 articles selected from the “WoS” database and employs three dimensions of technology, people, and institutions as a balanced perspective on smart cities, providing a comprehensive understanding of research trends on the “Metaverse” in the context of smart cities. This study identifies the “Metaverse” as a Virtual reality technology, popular since 2021, and provides information on the active years, countries, fields, journals, authors, and institutions involved in “Metaverse” research on smart cities. This study also identifies three stages of research development as follows: Stage 1 (2007–2013) to Stage 2 (2014–2020) and Stage 3 (2021–20 October 2023), revealing the research focus evolution from basic “urban planning” to complex “urban governance” and “Smart city” construction with consideration of multi-stakeholders’ perspectives. Additionally, this study reveals that “Metaverse” research studies on the “technology” dimension have consistently outnumbered that on “institutions” and “people” across all stages in the “Smart city” domain. These findings address current theoretical gaps and offer a foundation for future research.

1. Introduction

The phrase “Metaverse” is derived from the prefix “Meta” and the word “Verse”. The term “Meta” refers to “Transcendence”, whereas “Verse” relates to “Universe”. Both eventually allude to a parallel or virtual environment that is linked to the physical world [1,2,3]. The idea of the “Metaverse” is not brand-new as it has existed for more than 30 years since its first appearance in the science-fiction novel Snow Crash in 1992 [4]. The real-world application of the “Metaverse” often relies on immersive reality technologies such as Virtual reality (VR), Augmented reality (AR), Mixed reality (MR), Extended reality (XR), Digital twins (DTs), and avatars, but it extends beyond the spaces these technologies create [5]. The fourth industrial revolution and the changes brought about by the COVID-19 pandemic, including the formation of a “New normal” with more home-bound activities [6], have accelerated the development of other cutting-edge technologies. These advancements—5G/6G networks, artificial intelligence (AI), big data, the Internet of Things (IoT), 3D, edge computing, and blockchain—are rapidly evolving and seamlessly integrating the virtual and real worlds in the “Metaverse” [2,7].

The “Metaverse” is also considered the next evolutionary stage of the Internet [8], which allows users worldwide to interact socially and economically with the improvement of virtual experiences [3,9,10]. Along with consistent technological innovation, the rapid evolution of the “Metaverse” presents unprecedented opportunities for global interaction [11]. Gartner consultancy expects that by 2026, 25% of individuals will spend at least one hour daily in the “Metaverse” for work, shopping, education, socializing, or amusement [12]. Similarly, TT Consultants forecasts the XR market to grow rapidly at 22.7% on average per year by 2028, especially between 2023 and 2028, driven by the current intense enthusiasm and interest in this technology [13]. Additionally, another professional analysis agency, Mordor Intelligence, also predicts that the XR market will increase from USD 105.58 billion in 2023 to USD 472.39 billion by 2028 [14]. These international speedy surges in popularity highlight the transformative potential of the “Metaverse” in reshaping various industries [15,16]. Many sectors have been rapidly incorporating the “Metaverse” into their business to enhance service [17], typical examples can be found in game companies (e.g., Nintendo, NetEase, MiHoYo, and ZQGame), Internet giants (e.g., Google, Microsoft, Nivida, and Meta), and areas such as education [1,3,18,19], transportation [20,21], healthcare [22,23], and tourism [15,16,24,25]. In addition to the above domains, governments and public sectors have also started incorporating the “Metaverse” into municipal governance, particularly in the context of smart cities [7,16,17].

Smart cities are urban environments that leverage advanced digital technologies, infrastructure, systems, and innovative methods to optimize resource utilization, improve urban services, increase economic productivity, promote sustainable development, and enhance the quality of life for residents [26,27,28]. As urbanization intensifies, the demand for efficient, technologically advanced cities becomes more pressing, thus inspiring urban planners, developers, and administrators to concentrate on advancing the concept of the “Smart city”, while taking citizens’ welfare into account by emphasizing a variety of technological advancements along with human, environmental, social, cultural, and energy aspects [29,30]. In line with this trend, the introduction of the “Metaverse” into smart cities has become popular because it can offer transformative potential in enhancing citizen engagement, improving governance, and facilitating new forms of interaction between individuals and their urban environments [31].

However, the development of smart cities is not solely driven by technology. Smart cities should be understood as complex ecosystems where the interactions among residents, their environments, and societal factors play an equally vital role in shaping their growth and sustainability [32,33]. Accordingly, the concept of a “Smart city” should be approached from multiple dimensions, not only focusing on the part of technology. In this context, technology should be strategically integrated with other aspects of urban development to achieve smartness in the urbanization process, such as collaborative planning, behavior modification, community development, investment packages, multi-stakeholder engagement, and social, economic, and environmental issues [34,35]. As such, many researchers have introduced diverse models that account for these multifaceted interactions to capture the complexity of building a “Smart city”. For example, in 2007, Giffinger [36] introduced a six-dimensional “Smart city” model, focusing on economy, mobility, environment, people, living, and governance. Later, in 2011, Caragliu et al. [37] pointed out another six-dimensional “Smart city” model, involving urban infrastructure, business-driven urban development, high-tech industries, inclusive strategies, social and relational capital, and social and environmental sustainability. At the same time, Nam and Pardo [38] suggested a three-dimensional framework to highlight the roles of technology, people, and institutions in the development of smart cities. Recently, in 2019, Bolívar et al. [39] further defined a “Smart city” through six parts including information and communication technologies (ICTs), entrepreneurship, inclusiveness, social capital, and economic and environmental issues. Although these frameworks share common characteristics for conniving smart cities, the practical application in empirical studies still has limitations, particularly for those models with more dimensions. More dimensions mean that more corresponding evaluation metrics and standards are required to recognize the comprehensive development of specific urban functions and services. However, despite the increased number of metrics and standards, they often fall short in effectively evaluating the overall smartness of a city, face difficulties in integrating diverse technological and social dimensions, and struggle to ensure inclusivity and accessibility for all citizens [40]. Moreover, too many metrics and stands would also pose challenges to those cities with limited resources or at certain stages of development [41,42]. Thus, considering that the development of the “Metaverse” in a “Smart city” is still in its infancy stage [8], it appears to be more relevant and reliable to approach the “Smart city” from the three classical dimensions of technology, people, and institutions suggested by Nam and Pardo [38] because this model can offer adaptability and flexibility and can be tailored to the specific needs and developmental stages of different cities.

Nam and Pardo’s [38] three-dimensional theoretical framework emphasizes that the development of smart cities relies not only on technological advancements but also on the deep integration of social and institutional aspects. In the technology dimension, a “Smart city” prioritizes the use of ICTs to transform existing diverse services into more intelligent, virtual, informational, and ubiquitous forms, which aims to enhance the lives and work of citizens [38,43,44,45]. In the people dimension, a “Smart city” focuses on fostering creativity, diversity, and education to cultivate talented individuals capable of developing smart solutions to urban challenges [38], which ultimately leads to practical innovation and increased productivity in human and social capital, thereby enhancing smart living in cities [29,36,43,46,47]. In the institutions dimension, a “Smart city” emphasizes encouraging broader participation from individual citizen members, institutions, and other stakeholders (i.e., businesses, government agencies, and community organizations) in decision-making processes to enhance urban governance and growth [38].

Unlike other “Smart city” models that often include additional dimensions such as resource and environment as independent aspects to evaluate city development [36,37,39], Nam and Pardo’s [38] three-dimensional model inherently addresses these factors through both the people and institutions dimensions. From the perspective of people, smart cities cultivate individuals capable of creating smart solutions that address resource wastage and environmental challenges, with innovations in smart energy management, waste reduction technologies, and water conservation strategies, to optimize the use of natural resources and contribute to the sustainable urban ecosystem [38,43]. From the perspective of institutions, in the face of increasingly pressing environmental challenges, the collaborative efforts of multiple stakeholders become particularly important. Broad stakeholders, including government, businesses, and community organizations, work together to formulate policies and initiatives aimed at balancing technological progress with the sustainable development of smart cities. This collaboration not only helps address environmental issues but also promotes the long-term sustainable development of smart cities [48,49,50]. Although Nam and Pardo’s [38] model has fewer dimensions, it does not hinder a comprehensive evaluation of all parts of smart cities (e.g., technology, economy, resources, environment), as the success of building a “Smart city” largely depends on fostering collaborative relationships among all participants and maintaining a strong focus on citizen-centered efforts [43,48,49,50].

In addition, as the research on the “Metaverse” is still in its early stages [8], there are limited studies that address both the “Metaverse” and “Smart city” concepts simultaneously. Even where such studies exist, most of the related literature focuses primarily on technological aspects or specific domains, with little attention given to the intersection of these two fields. For example, Feng et al. [51], Shen et al. [52], Wider et al. [53], and Nan et al. [54] focused only on presenting key ideas and directions for developing “Metaverse” technology. These studies mainly emphasized the technical aspects of the “Metaverse” and did not address its potential applications and impacts in areas such as urban planning, infrastructure development, public services, and citizen engagement, especially in the context of smart cities. On the other hand, some researchers have recognized the importance of the “Metaverse” beyond its technical functions and argued for its integration into various areas of urban development. Nonetheless, their research generally tended to be limited to specific aspects of smart cities. For instance, Chen and Zhang [5] investigated the “Metaverse” development in the medical and health fields; Morais et al. [55] and Ha et al. [56] explored “Metaverse” applications in the fields of tourism, travel, and hospitality. Anshari et al. [57] examined the usage of the “Metaverse” by organizations in the business and commercial fields. Tas and Bolat [58], Bizel [59], and Chen et al. [60] analyzed “Metaverse” research in the education area. These researchers investigated the application of the “Metaverse” through their own literature analysis, but the target areas were limited to specific areas of interest within a “Smart city”, such as medicine, tourism, business, and education. In addition to the above studies, recently, a few other researchers have started to realize the research scarcity in this intersecting field and made contributions by publishing bibliometric research that simultaneously focuses on the “Metaverse” and a “Smart city”. Although these bibliometric studies are still rare, they have attempted to analyze the progress of related studies comprehensively, including research topics, publication trends, and implementation feasibility and limitations [6,17,30,61,62]. Among these, five representative review studies can be summarized, as shown in

Table 1. Five selected representative review-oriented studies covering the aspects of the “Metaverse” and “Smart city”.

No.	Source	Title	Methodology	Main Research Contribution	Understanding of the “Metaverse” Definition
1	[30]	The Metaverse as a Virtual Form of Smart Cities: Opportunities and Challenges for Environmental, Economic, and Social Sustainability in Urban Futures	literature review	Explored Metaverse-related potential contributions to smart cities, especially focusing on the aspects of the environment, economic, and social goals of sustainability.	The Metaverse as a virtual extension of smart cities, focusing on sustainability aspects including environmental, economic, and social dimensions.
2	[17]	Metaverse fundamental technologies for smart city: A literature review	systematic literature review	Provided a broader understanding of applying the Metaverse concept to smart cities and the underlying technologies that support it.	The Metaverse as an application of fundamental technologies in smart cities, emphasizing technological support and integration.
3	[61]	The Roadmap to Smart Cities: A Bibliometric Literature Review on Smart Cities’ Trends before and after the COVID-19 Pandemic	bibliographic analysis	Offered an intriguing grasp of the evolution of the Smart city concept in light of climate change, with the goal of detecting elements of innovation, emphasizing Metaverse implementation roadmaps and trends while also looking for modifications in research due to the COVID-19 pandemic.	The Metaverse as part of the smart city evolution, highlighting implementation roadmaps, trends, and the impact of COVID-19 on research directions.
4	[6]	The Metaverse as a virtual form of data-driven smart cities: the ethics of the hyper-connectivity, datafication, algorithmization, and platformization of urban society	literature survey	Investigated the Metaverse as a virtual version of data-driven smart cities, focusing on scopes of privacy, surveillance capitalism, dataveillance, geosurveillance, human health and wellness, and collective and cognitive echo-chambers.	The Metaverse as a virtual form of data-driven smart cities, addressing ethical issues such as privacy, surveillance, and health.
5	[62]	Metaverse applications in smart cities: Enabling technologies, opportunities, challenges, and future directions	literature review, project and case studies	Revealed the enabling technologies, opportunities, challenges, and future directions on the Metaverse implementation in smart cities.	The Metaverse as a platform for various applications in smart cities, focusing on enabling technologies, opportunities, and future directions.

.

By reviewing these five studies, this paper finds that the existing research on the “Metaverse” in the context of smart cities can provide some insights but lacks a unified definition of the “Metaverse”. In addition, the depth of analysis in these studies primarily remains at the level of describing technological and conceptual frameworks, with little attention given to the multidimensional aspects that are crucial for building smart cities. Moreover, given the recent surge in interest and publications related to the “Metaverse” in smart cities, these literature reviews do not sufficiently capture the latest research trends. This raises concerns about the up-to-dateness of research in the existing review studies.

Thus, to address the research gaps, this paper aims to identify the research status and trends in the “Metaverse” in the context of smart cities with multifaced views, especially based on the classical three-dimensional framework of a “Smart city” put forward by Nam and Pardo [38]. To conduct this research, this paper proposes the following research questions (RQs): (1) What is the “Metaverse”, and how is it being explored in the context of smart cities? and (2) Regarding the implementation of the “Metaverse” in the context of smart cities, what are the research trends, especially in the three dimensions of technology, people, and institutions? To find answers to these two RQs, bibliometric analysis is adopted in this paper. Bibliometric analysis involves using performance analysis and science mapping to examine the academic literature systematically and identify patterns, tendencies, and relationships within a specific field of study [63,64]. Through this approach, the analysis results and discussions presented in this study are expected to provide useful information on the topic. Compared with previous studies, which often focus on certain aspects of technology or specific dimensions, this research not only deepens the understanding of the intersection between the “Metaverse” and “Smart city”, but it also offers a more comprehensive and structured foundation—across the three dimensions of technology, people, and institutions—for future exploration by researchers in academia.

This paper is organized as follows. The Section 2 reviews the literature on the “Metaverse” and “Smart city” concepts. Section 3 introduces the bibliometric analysis methodology used in this study, focusing on the detailed data collection and research techniques. Section 4 performs bibliometric analysis to find answers to the two RQs presented earlier in this paper. Finally, Section 5 offers a discussion and conclusion.

2. Literature Review

2.1. Metaverse

The “Metaverse” is a technological term that has garnered significant attention in recent years, especially concerning a series of immersive reality technologies such as VR, AR, MR, and XR [1,2,3]. Even though the “Metaverse” is a concept that is over 30 years old, there is no universal definition for it. The understanding of the “Metaverse” has evolved significantly over time. Initially, the “Metaverse” was a concept based primarily on literature and speculative fiction [2]. However, with the development of immersive reality technology making it possible to build interactive digital environments, the potential applications and impact of the “Metaverse” have increased dramatically [65]. This evolution has led to the creation of opportunities for a wide range of experiences in the virtual world that rival those in the physical world [66]. According to Wang et al. [2]’s research, the development of the “Metaverse” can be divided into four stages, as shown in

Table 2. Development stages of the Metaverse.

Period	Stage	Description
1995–2006	Embryonic Stage	Mainly based on literature and art
2007–2013	Primary Stage	Mainly based on video games
2014–2020	Ebb Stage	More than just a game; starts to be contained in an economic structure and integrated into a human society
2021–2022	Development Stage	Refers to a series combination of new technologies to access multifield applications

Source: Re-arranged by authors based on Wang et al. (2023) [2].

.

The Embryonic Stage (1995–2006) saw most ideas about the “Metaverse” rooted in fiction novels and films, with literature and art as the primary conceptual carriers [2]. For example, the book Snow Crash [4] described the “Metaverse” as a virtual environment parallel to the real world, and the movie The Matrix (1999) portrayed a virtual universe that is controlled by an AI system. In the Primary Stage (2007–2013), video games became a key medium for early VR exploration, allowing players to create and customize avatars for interacting, shopping, and conducting business in immersive virtual environments [2,67]. The Linden Lab game “Second Life” is a typical example of this stage. During the Ebb Stage (2014–2020), the “Metaverse” expanded beyond gaming to include a more complex virtual economic structure, integrating deeper into human society [65]. This evolution aligns with Zuckerberg’s vision [68], who said, “VR products (e.g., Oculus) have the opportunity to create the most social platform ever and change the way we work, play, and communicate”. In the Development Stage (2021–2022), the “Metaverse” referred to a combination of a series of immersive reality and other emerging technologies. The “Metaverse” is seen as a hypothetical 3D network of virtual worlds that is distinguished by an immersive, transcendent, continuous, and empyrean cyberspace [66]. In such space, users can live in a very realistic way by interacting with various elements in the virtual world through their avatars, and this virtual-to-physical connection is not limited to visual and auditory experiences but involves deeper psychological and emotional bonds [66].

To bring the “Metaverse” to reality, it is widely recognized that immersive virtual technologies (e.g., VR, AR, MR, XR, DT, avatars) serve as the core drivers of its development [17,69,70,71,72]. In particular, VR creates an immersive virtual environment, allowing users to interact while being isolated from the physical world [17,71,73], whereas AR overlays virtual information onto the real world, blurring the line between virtual and physical spaces through digital images or information [71,72]. MR combines elements of both virtual and physical realities, enabling users to interact with objects in both environments simultaneously [30,71]. XR serves as an umbrella term encompassing VR, AR, and MR, which represents all forms of experiences that transcend physical reality [30]. A DT is a virtual replica of a physical system that allows real-time monitoring and optimization [30,74], while an avatar is a customizable digital representation of a user in the virtual world, used for socializing, working, and entertainment [2,17,30,66]. Aside from the aforementioned immersive virtual technologies, the reality of the “Metaverse” also needs to integrate other cutting-edge technologies. For example, blockchain technology strengthens the infrastructure of the “Metaverse”, particularly through the proliferation of non-fungible tokens (NFTs), which enables users to own, trade, and manage unique digital assets within the virtual world [17,30,31,75]. NFTs are transforming traditional economic models, creating emerging markets in virtual real estate, digital art, and virtual fashion [31]. In addition, 5G/6G networks and cloud computing technologies provide crucial support for the “Metaverse”, enabling real-time interactions and large-scale participation [17]. The low latency and high bandwidth of 5G/6G allow users to engage in complex virtual activities globally, from meetings to exhibitions and immersive tourism. Cloud computing ensures swift rendering and data processing in these virtual environments, enhancing the fluidity and realism of interactions [17,75]. In addition, the introduction of IoT technology allows users to interact within the “Metaverse” while remotely monitoring and controlling real-world devices [17,75]. All these technologies work to blur the boundaries between the virtual and physical worlds, which makes expanding “Metaverse” applications possible.

2.2. Smart City

Although the term “Smart city” was first used in the 1990s [29,46], it has now developed into an umbrella concept [40,76]. The word “smart” is fuzzy and inconsistent [46], and can be understood as “intelligent”, “digital”, “sustainable”, or “knowledgeable” [41,77,78,79], making it difficult to establish a unified definition of the “Smart city”. During the early period between the 1990s and 2000s, most studies on “Smart city” were generally rooted in the domains of sustainability, digital technology, or knowledge [43], forming it as a multidisciplinary concept [80]. Many researchers, academic institutions, governments, global companies, and international organizations have attempted to describe “Smart city” and provide their own insights [28,29,43,79,81,82]. According to these studies, the concept of a “Smart city” is linked to existing intelligent infrastructures, networks, and information, as well as ideals such as accountability, collaboration, and participation. Despite these efforts, defining a “Smart city“ remains a challenge because of the diverse ways in which the concept is applied globally [38,83]. While the term “Smart city” is widely recognized, it is often referred to by different names in various regions and contexts, leading to a range of conceptual variations. For example, “Smart city” focuses on the integration of technology into urban systems, “Digital city” emphasizes the use of ICTs for communication and services, “Intelligent city” centers around AI-driven decision-making [79,84,85], and “Cognitive city” represents cities that are adaptive and capable of learning from data [83]. These variants are often created by substituting “smart” with various adjectives, which further complicates the development of a universal definition of the “Smart city” [38,83]. Thus, to address these complexities and establish a common understanding of a “Smart city”, it is essential to examine the closely related concepts from many dimensions. Nam and Pardo [38] provided the most commonly known model to understand a “Smart city” from the three dimensions of technology, people, and institutions.

Table 3. Related implications of a “Smart city” under the dimensions of technology, people, and institutions.

Dimensions	Related Implications	Selected Representative Source	Proof Quotes (Exemplary)
Technology	Digital City; Virtual City, Information City; Wired City; Ubiquitous City; Intelligent City	[45,78,86,87]	“Smart city is defined by IBM as the use of information and communication technology to sense, analyze and integrate the key information of core systems in running cities” [86].
People	Learning City; Knowledge City; Creative City; Cognitive city; Human City	[37,46,47,88]	“People with higher education and creative planning systems are required, in order to have a smart city with high productivity” [88].
Institution	Smart City; Smart Community; Sustainable City; Green City; Collaborative City; Eco City	[89,90,91,92]	“The concept of a smart community refers to the use of information and communication technologies by local governments and cities to better interact with their citizens, taking advantage of all available data to solve important problems” [91].

below outlines the implications of a “Smart city” under Nam and Pardo’s [38] three-dimensional theoretical framework, with selected representative proof quotes. In the technology dimension, the term “smart” can be replaced by “digital”, “virtual”, “information”, “wired”, “ubiquitous” and “intelligent”, which stress the focus on the integration of digital and intelligent infrastructures [45,78,86,87]. In the people dimension, the term “smart” can be understood as “learning”, “knowledge”, “creative”, “cognitive” and “human”, which emphasizes the importance of human capital and social engagement [37,46,47,88]. In the institution dimension, the term “smart” can be synthesized as ”smart community”, “sustainable”, “collaborative”, “green” and “eco”, which highlight the role of governance and institutional collaboration in the development of smart urban environments [89,90,91,92].

With the continuous development of the concept, the definition and connotation of the “Smart city” have gradually evolved and expanded. By 2018, the extensive research on the “Smart city” had transformed it into an independent research discipline [43,93], shifting the research focus from primarily ICTs and technological infrastructures to improving citizens’ well-being and promoting sustainable development. In 2018, the International Telecommunications Union (ITU) [92] also offered the most comprehensive definition of a “Smart city” as follows: “A smart sustainable city is an innovative city that uses ICTs and other means to improve quality of life, efficiency of urban operation and services, and competitiveness, while ensuring that it meets the needs of present and future generations with respect to economic, social, environmental as well as cultural aspects”. This definition has been accepted by over 300 experts worldwide and reflects the six most common known components of a “Smart city”—smart economy, smart people, smart governance, smart mobility, smart environment, and smart living [29,36,41,42]. In greater detail, smart economy refers to the integration of digital technologies to enhance economic competitiveness and innovation. Smart people emphasize the role of human capital and citizen engagement in fostering a knowledge-driven and inclusive society. Smart governance involves the use of ICTs to improve public services and decision-making processes. Smart mobility addresses the application of advanced technologies in transportation systems to ensure efficient, sustainable, and safe mobility within urban environments. Smart environment focuses on sustainable resource management and reducing environmental impacts through the use of ICTs. Smart living encompasses efforts to improve quality of life, including healthcare, education, safety, and housing, by integrating technology into everyday life.

2.3. The Metaverse for the Smart City

The integration of the “Metaverse” into urban development is gaining significant attention as cities seek innovative strategies to comprehensively address concerns related to the environment, production, and economics [30]. By using a series of immersive and interactive technologies like VR, AR, MR, XR, AI, blockchain, NFTs, and cloud computing, the “Metaverse” holds the potential to revolutionize city planning and service delivery, improving urban efficiencies, accountability, and the quality of public services [30,31,94].

In practice, the “Metaverse”, particularly with its technologies like DTs and avatars, offers new opportunities for city planners to enhance preparedness capacities to address “what-if” scenarios through modeling and simulating events like natural disasters (e.g., floods, earthquakes), public safety incidents, energy demands, climate change, population migration, and environmental changes [6,7,30,66]. Beyond its application in city planning, the “Metaverse” offers significant opportunities in the realm of urban governance. By adopting immersive technologies (e.g., VR, AR, MR, XR) and utilizing real-time analysis through “Metaverse” platforms, local governments can enhance interactions with residents, deliver fast and efficient real-time services, and optimize the management of urban assets such as public spaces [30,95]. Cities like Seoul, Orlando, and Las Vegas are the forefront representatives of integrating “Metaverse” technologies to achieve smarter, more connected urban environments with improved urban governance [17,62,65]. With the use of VR or AR equipment, residents from these cities can gain immersive experiences and access government services such as filing complaints [96]. In addition, the “Metaverse” has the potential to transform the economy and business by boosting brand visibility in digital marketplaces, making production more efficient, simplifying inventory management, and helping virtual economies grow with advanced cryptocurrencies [69,70]. Furthermore, “Metaverse” technologies can support transportation systems by integrating virtual simulations for traffic control, autonomous vehicle management, and route optimization, which improve the efficiency and sustainability of urban mobility [6,7,30]. In addition, the “Metaverse” can reduce resource consumption by replacing physical products with virtual alternatives, particularly items that are rarely used, thereby minimizing waste and lowering energy demand, which in turn reduces global emissions [30]. To further support decarbonization efforts, the “Metaverse” also reduces the need for travel through virtual workspaces and aids environmental protection by enabling real-time monitoring and optimization of resources like water and energy, thereby promoting more sustainable practices and advancing broader climate goals [30].

However, despite the great potential of the “Metaverse” to enhance governance and interaction efficiency in cities, its implementation relies on a range of enabling technologies that demand high computational power, which can significantly increase energy consumption and negatively impact environmental protection. Schomaker et al. [97] mentioned that while data centers are crucial for modern information technologies (e.g., the “Metaverse”), the increasing demand for cloud services and web applications would result in higher use of water and electricity resources, not only by IT hardware but also by supporting infrastructure like power supply and cooling systems. Garnett [98] and Babel et al. [99], further emphasized the energy-intensive nature of the virtual economy in the “Metaverse”, particularly highlighting the heavy energy demands of blockchain technologies like NFTs and cryptocurrencies. The Proof of Work (PoW) method, commonly used in public blockchains for creating NFTs, consumes vast amounts of electricity, which contributes significantly to carbon dioxide emissions and exacerbates climate change. Graham [100] reported that by 2025, AI’s energy consumption may surpass that of the human workforce unless sustainable practices are adopted, which could undermine progress toward achieving net-zero carbon goals. Recently, Nleya and Velempini [101] pointed out that even though the “Metaverse” has the potential for both positive and negative environmental impacts, its increasing energy consumption and e-waste generation are concerns. All these reveal that the “Metaverse” and its enabling technologies have high computational demands, which eventually and inevitably leads to heavy resource use and creates a conflict with the sustainability goals of smart cities. This has also evolved into an open issue, compelling future research to address how the “Metaverse” can be implemented in a way that balances its potential benefits with the need to minimize resource consumption and support long-term sustainability.

3. Methodology

The primary research method used in this paper is a bibliometric analysis. As mentioned by Moral-Muñoz et al. [63] and Donthu et al. [64], bibliometric analysis is a common and thorough method for exploring and interpreting vast amounts of scientific information, allowing researchers to discover the evolutionary aspects of a particular discipline while also offering insights into the growing areas in that field. Researchers employ bibliographic analysis for a variety of purposes, including examining publication and journal performance, collaboration patterns, trends in the development of research aspects, and conceptual frameworks used in the existing literature [64,102]. Regarding bibliometric analysis, Donthu et al. [64] outlined the process in the following four stages: (1) define the aims and scope of the bibliometric study; (2) choose the techniques for bibliometric analysis; (3) collect the data for bibliometric analysis; and (4) run the bibliometric analysis and report the findings. These stages provide a standardized procedure for bibliometric studies, so this study follows them.

In this paper’s context, the objectives and scope of the bibliometric analysis in relation to the first stage were presented in Section 1 by defining two RQs. For the second stage, this study selected the following main techniques that are well-established in specific procedures and used in bibliometric studies: (1) performance analysis, which evaluates the productivity and impact of researchers, institutions, and countries, and (2) science mapping, which visualizes the structural and dynamic aspects of scientific research [63,64]. The third stage relates to collecting data (literature) by using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) tool [103]. The PRISMA tool is crucial for conducting literature reviews systematically, as it ensures that the data (literature) selection process is systematic, transparent, and reproducible through clear processes and standards [103]. The fourth stage involves the actual execution of bibliometric analysis, where the analysis results, research findings, discussion, and conclusion are primarily presented. The corresponding procedures of these four stages, as guided by Donthu et al. [64] and simultaneously integrated with the PRISMA tool, are shown in Figure 1.

3.1. Data (Literature) Collection

The database used for searching in this step is the “Web of Science (WoS)”, which is widely recognized as the world’s most reliable and publisher-independent reference source. “WoS” provides access to numerous databases and citation information across various fields, including science and social science; it contains over 90 million records and covers publications from 1900 to the present [63]. “WoS” is also recognized as a robust literature database for conducting bibliometric analysis [63,64].

In this step, the keyword string used for searching is firstly analyzed by setting inclusion and exclusion criteria. Given the study objective is the intersection between the “Metaverse” and “Smart city”, the inclusion criteria are that the keyword string for searching simultaneously covers resources related to both the “Metaverse” and “Smart city”. In contrast, the exclusion criteria are set to exclude resources that deal only with either the “Metaverse” or “Smart city”. Based on both the “Metaverse” and “Smart city” concepts discussed in Section 2, the detailed research strings are listed in

Table 4. Research strings for keywords.

Code	Descriptions	Research Strings
Key1	“Metaverse”-related searching	“Metaverse” OR “immersive technolog*” OR “Virtual reality” OR “VR” OR “Mixed reality” OR “MR” OR “Augmented reality” OR “AR” OR “Extended reality” OR “XR” OR “Digital twins” OR “DT” OR “avatar” OR “Virtual space” OR “virtual environment” OR “virtual world”
Key2	“Smart city”-related searching	“Smart cit*” OR “Meta cit*” OR “Digital cit*” OR “Intelligent cit*” OR “Knowledge cit*” OR “Sustainable cit*” OR “Smart urban*” OR “Meta urban*” OR “Digital urban*” OR “Intelligent urban*” OR “Knowledge urban*” OR “Sustainable urban*” OR “Smart communit*” OR “Meta communit*” OR “Digital communit*” OR “Intelligent communit*” OR “Knowledge communit*” OR “Sustainable communit*” OR “Smart government” OR “Meta government” OR “Digital government” OR “Intelligent government” OR “Knowledge government” OR “Sustainable government” OR “Smart municipal” OR “Meta municipal” OR “Digital municipal” OR “Intelligent municipal” OR “Knowledge municipal” OR “Sustainable municipal”
Key3	Searching simultaneously related to the scopes of the “Metaverse” and “Smart city”	#Key1# And #Key2#

Note: The symbol “*” indicates one or more characters in search strings, and the symbol “#” is used to distinguish specific keyword groups.

.

The code “Key1” is designated for search terms related to the “Metaverse”. It includes research strings that capture a broad range of virtual environments and associated technologies, as outlined in the “Metaverse” definition review in Section 2.1. These terms encompass various immersive and virtual technologies, such as “Metaverse”, “Virtual reality (VR)”, “Mixed reality (MR)”, “Augmented reality (AR)”, “Extended reality (XR)”, “Digital twins (DTs)”, “avatar”, “Virtual space”, “virtual environment”, and “virtual world”. To ensure comprehensive coverage of these concepts, the search string for “Key1” is constructed using the Boolean operator “OR”, allowing for the inclusion of all relevant terms under the “Metaverse” domain.

The code “Key2” is used to search for terms related to “Smart city”. Based on the conceptual review in Section 2.2, the research string incorporates the three-dimensional “Smart city” model of technology, people, and institutions, as proposed by Nam and Pardo [38]. Specifically, adjectives such as “smart”, “meta”, “digital”, and “intelligent” reflect the technological dimension, while “knowledge” and “sustainable” pertain to the people and institutional dimensions, respectively. Additionally, given different spellings, synonyms, or stem variations, strings like “cit*”, “urban”, “communit*”, “government”, and “municipal” are included to represent the urban context within a city comprehensively. The query for “Key2” is also constructed by using the Boolean operator “OR” to ensure all relevant terms are contained.

The code “Key3” represents the eventual search consideration, which is used by Boolean operators “AND” to combine code “Key1” and “Key2”. All the queries for “Key1”, “Key2”, and “Key3” are applied through the advanced search method in the “WoS” database. This method ensures that only documents addressing the scopes of both the “Metaverse” and “Smart city” are retrieved, thereby ensuring that the search meets the inclusion criteria.

To further ensure the appropriateness of the keyword string selection, beyond considering the inclusion and exclusion criteria when setting the search parameters, each author of this paper also carefully reviewed and confirmed that the search strings accurately cover both the “Metaverse” and “Smart city” aspects, while fully addressing the three dimensions of technology, people, and institutions within the context of smart cities. This review and confirmation process improved the precision of the search and helped to avoid potentially misleading results.

In addition, this paper further refined the search criteria by specifying additional conditions to ensure the relevance and quality of the search results. The additional conditions limit the data (literature) to being published in English, including articles, review articles, and conference proceedings, and also limit the publication date range from 2007 to 2023 (ending on 20 October). Moreover, this paper considers only articles indexed in the Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Arts and Humanities Citation Index (A&HCI), and Conference Proceedings Citation Index (CPCI) to ensure the selection of high-quality documents.

To perform the data (literature) collection process effectively, this study utilized the PRISM flow chart [103], which is illustrated in Figure 2. As shown, this study identified 604 papers from “WoS” for further bibliometric analysis, which include both performance analysis and science mapping techniques. The number of selected documents exceeding 500 indicates its suitability for conducting bibliographic analysis [64].

3.2. Performance Analysis

Performance analysis, as one of the leading research techniques of bibliometric analysis, explores the contributions of research units to a specific field by considering metrics such as publication counts and citation frequencies [63,64]. By analyzing these metrics, performance analysis can effectively identify relevant research trends while allowing researchers to identify major contributors and influential works in the field [63,64].

3.3. Science Mapping

Science mapping, another key technique of bibliometric analysis, investigates the relationships among research components [63,64,104]. This technique helps uncover patterns, trends, and connections among different research topics, disciplines, and collaborations. Science mapping can be further enhanced by integration with the social network analysis (SNA) tool, which allows for a more detailed analysis of the research trends in the study field [63,64,104]. This paper uses the SNA software “Netminer 4” (Version 4.5.1.a) for science mapping, especially for keyword analysis and topic modeling analysis.

4. Bibliometric Analysis Results

4.1. Performance Analysis Results

This section presents a comprehensive performance analysis, focusing on how “Metaverse”-related studies were conducted in the context of smart cities from the following six aspects: the distribution of published documents by year, the distribution of productive countries, the distribution of research fields, the distribution of productive sources, the distribution of authors, and the distribution of affiliations.

• Distribution of Published Documents by Year.

The trend graph (Figure 3) below, obtained from the “WoS” database, shows that research on the “Metaverse” in the context of smart cities began to appear as a notable theme in 2007. The steepest growth trend is observed after 2014. The number of related papers published increased from 1 in 2007 to 135 in 2022 and 89 as of 20 October 2023.

• Distribution of Productive Countries.

Examining the distribution of documents by country, it was found that nearly 79 countries and regions contributed to the research on the “Metaverse” in the context of smart cities. The top 15 countries are China (144, 23.68%), USA (77, 12.66%), South Korea (47, 7.73%), England (44, 7.24%), Spain (41, 6.74%), Italy (39, 6.41%), Germany (35, 5.76%), Australia (30, 4.93%), India (28, 4.61%), Saudi Arabia (27, 4.44%), Sweden (21, 3.45%), France (18, 2.96%), Canada (15, 2.51%), Romania (15, 2.47%), and Greece (14, 2.30%). Figure 4 illustrates the number of records from these countries.

• Distribution of Research Fields.

Regarding the research fields, the majority of the publications are from the field of “Computer Science (51.15%)”, followed by “Engineering (38.49%)”, “Telecommunications (19.08%)”, “Environmental Sciences Ecology (11.68%)”, “Science Technology Other Topics (10.53%)”, “Remote Sensing (6.25%)”, “Imaging Science Photographic Technology (5.26%)”, “Chemistry (4.61%)”, “Construction Building Technology (3.95%)”, “Physical Geography (3.45%)”, “Business Economics (3.29%)”, and “Instruments Instrumentation (3.29%)”. Figure 5 illustrates the research areas with more than 20 records.

• Distribution of Productive Sources.

Concerning the productive sources, we found that “Sustainability (38, 6.25%)”, “IEEE Access (26, 4.28%)”, “Lecture Notes in Computer Science (22, 3.62%)”, and “Sensors (18, 2.96%)” showed high productivity. The following journals were also identified: “Applied Sciences Basel”, “Future Generation Computer Systems the International Journal of Escience”, “IEEE Internet of Things Journal”, “Sustainable Cities and Society”, “Remote Sensing”, “Buildings”, “Journal of Management in Engineering”, and “ISPRS International Journal of GEO Information”. Each of these journals published fewer than ten papers from 2007 to 20 October 2023 (Figure 6).

• Distribution of Productive Authors.

In terms of author distribution, 2111 individuals contributed to the “Metaverse” research centered on smart cities. Figure 7 lists the most productive authors with three or more works as follows: “Lv ZH (14)”, “Li XM (6)”, “Wang WX (6)”, “Svítek M (5)”, “Yang B (4)”, “Taylor JE (4)”, “Postránecky M (4)”, “Mohammadi N (4)”, “Liu Z (4)”, “Khan S (4)”, and “Chen Y (4)”.

• Distribution of Productive Affiliations.

Turning to affiliation distribution (Figure 8), the most productive institutions (with more than five works) were identified as following: “Chinese Academy of Sciences (15)”, “University System of Georgia (10)”, “Uppsala University (10)”, “Egyptian Knowledge Bank EKB (9)”, “Deakin University (8)”, “Beijing University of Posts Telecommunications (7)”, “Centre National De la Recherche Scientifique Cnrs (7)”, “Wuhan University (7)”, and “Qingdao University (6)”.

4.2. Science Mapping Results

In this study, a science mapping analysis was also conducted to visualize the systematic development status of “Metaverse” research in the field of smart cities across the three dimensions of technology, people, and institutions. This technique includes both keyword analysis and topic modeling analysis. Keyword analysis covers high-frequency keyword analysis and keyword network mapping, while topic modeling analysis includes topic extraction analysis and topic ratio analysis.

4.2.1. Keyword Analysis

• High-Frequency Keyword Analysis.

High-frequency keyword analysis focuses on identifying and examining the most frequently occurring keywords within a set of documents. This method often uses the word cloud analysis approach [51,105]. In this paper, Figure 9 is the word cloud that shows the most frequent keywords among selected sources about the “Metaverse” research related to smart cities. The size of each keyword label corresponds to its frequency of occurrence in the documents, with larger labels indicating higher frequencies and smaller labels indicating lower frequencies. Recognizing the frequency of keyword occurrence is crucial in achieving our research objectives because it helps identify a field’s hotspots and predict future research trends [51,105]. In Figure 9, it is evident that “Smart city”, “Digital twin”, “IoT”, “Augmented reality”, and “Technology”, are the top five keywords with the highest frequencies. These frequencies suggest that these keywords play an important role in the “Metaverse” research related to the “Smart city”.

• Keyword Network Mapping.

Keyword network mapping is a visualization analysis that displays the frequency and relationships among different keywords co-occurring in the literature or text data. By conducting keyword network mapping, researchers can visually observe the strength and structure of connections among various themes within a research field [106,107]. This study examined the concepts of the “Metaverse” and “Smart city” individually and found some similarities in their historical development. That is, both terms were first used in the 1990s, and they developed slowly until the late 2000s; however, with the emergence of new technological developments such as VR, AR, MR, XR, blockchain, and NFTs, both the “Metaverse” and “Smart city” experienced significant growth in the 2010s. To provide a structured understanding of “Metaverse” growth, Wang et al. [2] divided the development of the “Metaverse” into four stages, illustrated in Table 2, as follows: the Embryonic Stage (1995–2006), the Primary Stage (2007–2013), the Ebb Stage (2014–2020), and the Development Stage (2021–2022). This classification focuses solely on the technological aspects of the “Metaverse”. Thus, to reflect the development of the “Metaverse” research within the “Smart city” domain, this paper considered the similar development timelines of the “Metaverse” and “Smart city” and proposed dividing this research into three stages as follows: Stage 1 (2007–2013), Stage 2 (2014–2020), and Stage 3 (2021–October 20, 2023). This classification builds on the stages proposed by Wang et al. [2] but adapts them to capture the specific developmental milestones relevant to utilizing the “Metaverse” in the context of smart cities.

According to this classification, this paper constructed a keyword network for each stage, where keywords were treated as nodes. To gain deeper insights into the structure of these networks, an analysis was conducted using indices such as degree centrality and betweenness centrality associated with the nodes. Degree centrality was computed as the number of direct linkages to a node. A high degree centrality score indicates that a node has many linkages pointing to it, reflecting its potential importance in the research area [108]. Betweenness centrality was calculated based on the number of shortest paths from all vertices to all other nodes, which can also be considered a measure of a node’s importance. A node with high betweenness centrality can link key nodes to others, highlighting their potential importance as a research topic [108,109].

• Stage 1 (2007–2013).

In Stage 1, the number of relevant studies covering aspects of both the “Metaverse” and “Smart city” was quite limited, with only 26 papers. The keywords derived from these papers, along with their degree and betweenness centrality results, are presented in

Table 5. Degree and betweenness centrality of selected keywords in the three stages.

No.	Keywords	Degree Centrality	Betweenness Centrality	No.	Keywords	Degree Centrality	Betweenness Centrality	No.	Keywords	Degree Centrality	Betweenness Centrality	No.	Keywords	Degree Centrality	Betweenness Centrality
Stage 1 (2007–2013)				5	Virtual reality	0.0984	0.1604	24	CityGML	0.0328	0.0492	12	Competition	0.0698	0.0093
1	Urban planning	0.1667	0.0152	6	Brasov	0.0874	0.091	25	M2M	0.0328	0.0318	13	Distance learning	0.0698	0.0616
2	GIS	0.0833	0.0000	7	Mixed reality	0.0765	0.0894	26	Secure CoAP	0.0273	0.0359	14	LiDAR	0.0698	0.0240
3	Los Angeles	0.0833	0.0000	8	Innovation	0.0710	0.0504	27	Smart Cyber Physical Systems	0.0273	0.0112	15	Climate change	0.0581	0.0123
4	Metacognition	0.0833	0.0000	9	6G	0.0656	0.0744	28	Smart objects	0.0273	0.0039	16	Federated learning	0.0581	0.0108
5	Science and Engineering Design	0.0833	0.0000	10	Earth observation	0.0546	0.0448	29	Urbanization	0.0273	0.0052	17	Networking	0.0581	0.0141
6	Shanghai	0.0833	0.0000	11	Cycling	0.0546	0.0312	30	Competitiveness	0.0273	0.0021	18	Pandemic	0.0581	0.0248
7	Digital city	0.0833	0.0000	12	Simulation	0.0546	0.0612	Stage 3 (2021–20 October 2023)				19	Partnership	0.0581	0.0517
8	Infrastructure	0.0833	0.0000	13	Enterprise architecture framework	0.0546	0.0456	1	Internet	0.2442	0.2252	20	Path planning	0.0581	0.0099
9	Polymorphism computation	0.0833	0.0000	14	Interactive storytelling	0.0546	0.0679	2	Culture	0.2093	0.2255	21	DBSCAN	0.0465	0.0093
10	Simulation	0.0833	0.0000	15	Sensing	0.0492	0.0335	3	Emergency evacuation	0.1744	0.1421	22	Life cycle	0.0465	0.0216
11	Virtual city	0.0833	0.0000	16	VEoT	0.0492	0.0492	4	Food	0.1744	0.1443	23	Metaverse	0.0465	0.0215
12	Visualization interaction	0.0833	0.0000	17	City model	0.0437	0.0334	5	Inventory management	0.1628	0.1200	24	Network lifetime	0.0465	0.0012
13	Wake interference regime	0.0833	0.0000	18	Cognitive radio	0.0437	0.0309	6	Embodiment	0.1279	0.0465	25	Auralization	0.0349	0.0097
Stage 2 (2014–2020)				19	Data fusion	0.0383	0.0152	7	BIM	0.1163	0.0700	26	Blockchain	0.0349	0.0208
1	Urban governance	0.1694	0.2082	20	Economy	0.0383	0.0117	8	Brain tumor	0.1163	0.0374	27	Data driven	0.0349	0.0208
2	Characteristics	0.1257	0.1683	21	Governance	0.0383	0.0317	9	Caching	0.1163	0.0693	28	Data spaces	0.0349	0.0215
3	Laser diodes	0.1202	0.1281	22	Modularity	0.0328	0.0165	10	Digital citizenship	0.1047	0.0947	29	Decision support	0.0349	0.0227
4	Latency	0.1093	0.1083	23	Photogrammetry	0.0328	0.0176	11	Avatars	0.0930	0.0357	30	Digital transformation	0.0349	0.0211

, and their corresponding visualization is shown in Figure 10. The figure shows that during Stage 1, the keywords “urban planning”, “GIS”, “Metacognition”, “Science and Engineering Design”, “infrastructure”, “Polymorphic computation”, “simulation”, “Visualization interaction”, and “Wake interference regime” were central in the network, along with two detailed city names, “Los Angeles” and “Shanghai”. In particular, “urban planning” had the highest betweenness centrality, indicating that this keyword was an important mediator in the network connecting various topics and played an important role as the main focus of most of the research conducted at this stage. This is further evidenced by the papers published during this period. For instance, Liu et al. [110] emphasized that urban planning and management, urban modeling, and digital city construction are important activities to promote modern urban infrastructure. Dong et al. [111], Chen and Wu [112], and Liu and Peng [113] emphasized the need to utilize VR technology and earth science information in urban planning to create virtual cities. These studies show that early-stage research was primarily conducted from the perspective of designing basic urban infrastructure and suggested the need to enhance the construction of a “Digital city” or a “Virtual city” by utilizing VR-related technologies in urban planning.

• Stage 2 (2014–2020).

During Stage 2, the number of papers on “Metaverse” research related to smart cities increased to 284. From the keywords retrieved for these papers and the calculation results of each keyword’s degree and betweenness centrality, it seems that researchers showed a strong interest in technology during this period. Keywords like “laser diodes”, “latency”, “augmented reality”, “mixed reality”, “innovation”, “6G”, “simulation”, “interactive storytelling”, “sensing”, “VEoT” (the Virtual Environment of Things), “data fusion”, “modularity”, “photogrammetry”, “CityGML” (City Geography Markup Language), “M2M” (Machine to Machine), “Secure CoAP” (Secure Constrained Application Protocol), “smart cyber–physical systems”, and “smart objects” illustrated high degree centrality values. In addition, “urban governance” and “characteristics” exhibited the highest betweenness centrality during this period. To further highlight these influential topics during this period, this paper mainly focused on the top 30 keywords ranked by degree centrality while also considering betweenness centrality. Table 5 lists these keywords, and Figure 11 presents the corresponding keywords network in Stage 2.

Compared with Stage 1, the keyword occurrence results of Stage 2 show a greater proportion of those related to the potential of technology. The research topics shifted from early urban planning to more complex aspects of urban administration and governance, aiming to realize the ideal city model. These findings can also be seen in many papers published during this period. For example, Jutraz and Zupancic [114] presented a series of virtual tool systems in their study called “Terf”, emphasizing that it would be useful for designing, coordinating, constructing, and maintaining a “Smart city”. Portela and Granell Canut [115] emphasized that companies and governments need to enhance data processing capabilities to technologically address major issues faced by cities and improve their strategic urban management capabilities. Ahmed and Rani [116] regarded IoT technology as playing a major role in providing sustainable growth of digital services and dynamic realization of new city updates, which also would aid in providing value-added services to citizens and improving urban management performance in the future. Bellini et al. [117] pointed out the need to utilize technologies such as blockchain and VR in the operation of a “Smart city” to achieve high performance in urban governance. Dembski et al. [118] also stressed that digital technologies and their applications are essential to building smarter and more sustainable governance for democratic and intelligent urban development.

• Stage 3 (2021–20 October 2023).

For the most recent stage, beginning in 2021, a total of 294 papers were collected by the research cut-off date (20 October 2023). Based on these papers, this paper calculated the degree and betweenness centrality of keywords and found that research topics during this period became more diverse and multidimensional than in the previous stage. Technologies such as the “Internet”, “networking”, “blockchain”, “data-driven approaches”, “data spaces”, and “digital transformation” emerged prominently. In addition, new terms related to virtual technologies, such as “avatar” and the “Metaverse”, began to dominate the keyword network. The research focus appeared to shift towards covering “climate change” and “pandemic”. Other keywords such as “culture”, “emergency evacuation”, “food”, “inventory management”, “BIM” (Building Information Modeling), “brain tumor”, “life cycle”, and “distance learning”, also indicating that research on the “Metaverse” expanded to multiple aspects of a “Smart city”, encompassing fields such as supply chain, manufacturing, construction, transportation, healthcare, and education. The top 30 keywords in this stage are presented in Table 5, which are ranked primarily by degree centrality, with betweenness centrality also considered. The corresponding keyword network is shown in Figure 12.

Compared with Stage 2, the “Metaverse” research related to smart cities in Stage 3 was characterized by a rapid expansion of research volume, broader interdisciplinary penetration, and a stronger technological focus. In particular, terms such as “digital citizenship” and “partnership” achieved high betweenness centrality values, indicating a growing awareness of multi-stakeholder participation in Smart city construction. This shift reflects an increasing recognition of the importance of managing urban data to develop data-driven smart solutions and diversifying stakeholders in the construction of smart cities. Looking back at the research papers published in this stage, several studies also emphasized these aspects. Specifically, Petrovic et al. [119] stressed the importance of urban data management for developing data-driven smart solutions, building a sustainable Smart city ecosystem, and improving citizen well-being. This emphasis contrasts with Stage 2, which focused more on improving data processing capabilities, as pointed out by Portela and Granell-Canut [115]. That is, the research focus in Stage 3 placed greater emphasis on the deeper integration of technology, highlighting the shift from merely processing data to managing and leveraging it effectively to drive innovative Smart city solutions. In addition, Hunter et al. [120] argued that technological advancements and digital transformation can revolutionize local services and economies, fostering citizen connectivity and achieving sustainable outcomes for the environment, society, and urban areas. Similarly, Oschinsky et al. [121] found that creative collaboration techniques are essential for promoting smart city governance with a public-centric approach. Moreover, White et al. [122] highlighted that citizens’ roles should be prioritized when creating a DT-based 3D city model, as it allows citizens to engage with the model through various feedback mechanisms based on their own perspectives. In addition, Caputo et al. [123] proposed a sustainability-based conceptual framework to determine the best actions to increase citizen engagement and ensure sustainable growth in smart cities.

4.2.2. Topic Modeling Analysis

• Topic Extracted Analysis.

Topic modeling is a type of text-mining technique that has been widely applied in academia to explore hidden topics in a set of documents [124,125]. One of the representative algorithms is Latent Dirichlet Allocation (LDA), a generative probabilistic model that computes a given number of topics by considering the probability distribution of terms associated with each topic [126]. LDA-based topic modeling can extract topics by examining the word distribution across the documents, which is quite useful in identifying knowledge structures and patterns in research publications [127]. Accordingly, based on the selected 604 papers, this paper performed topic modeling using the LDA algorithm with the software “Netminer 4” (Version 4.5.1.a), and as a result, 10 topics were extracted, as shown in Figure 13.

Combining the word networks generated under each topic, this paper discovered that “Topic 1”, “Topic 4”, “Topic 5”, “Topic 8”, “Topic 9”, and “Topic 10” focus more on the technical aspects of exploring the application of virtual technologies in the “Smart city” domain. Notably, the newly popular terms “Digital twin” and “Metaverse” appear most frequently in “Topic 1”, “Topic 5”, and “Topic 9”. These three topics show the most recent developments in the field of the “Metaverse”, also known as VR technology. Meanwhile, “Topic 4”, “Topic 8”, and “Topic 10”, consist of high-probability terms such as “IoT”, “device”, “edge computing”, “Augmented reality”, “Virtual reality”, and “Machine learning”, focus on the fundamental technologies for realizing the “Metaverse”. Regarding the remaining topics, “Topic 2” is linked to the construction of smart cites, specifically supported by technologies related to digital modeling and simulation. This can be confirmed by the high-probability terms such as “image”, “scene”, “visualization”, “3D” (city models), “algorithms”, “navigation” and “urban planning”. These terms indicate that digital technologies are used to create and manage virtual representations of the urban environment, which is essential for the development of modern smart cities. “Topic 3” is associated with sustainable city construction, which focuses on urban planning and city governance that consider the perspectives of “community” and “citizens”. Representative terms for “Topic 3” include “innovation”, “impact”, “solution”, “tool”, and “implementation”. These terms highlight the importance of innovative solutions and tools in addressing the needs and impacts on communities and citizens in sustainable urban development. “Topic 6” is identified as focusing on education, supported by frequent terms such as “design”, “game”, “Virtual reality”, “student”, “learning”, “education”, and “knowledge”. These terms refer to the use of interactive and immersive technologies in educational settings to enhance student learning experiences. “Topic 7” is categorized as relating to tourism scenarios, indicated by key terms such as “user”, “tourism”, “place”, “experience”, “project”, “person”, and “culture”. These terms highlight enhancing user experiences in tourism through innovative projects focused on cultural and personal interactions.

Referring back to the three dimensions of a “Smart city”, the ten topics can be grouped accordingly into these categories. “Topic 1”, “Topic 4”, “Topic 5”, “Topic 8”, “Topic 9”, and “Topic 10” emphasize the technical aspects of applying virtual technologies, thereby fitting into the “technology” dimension. “Topic 6” and “Topic 7” concentrate on enhancing higher education and improving user experiences through various creative systems and projects, thus fitting into the “people” dimension. Meanwhile, “Topic 2” and “Topic 3” focus on building smart and sustainable cities while recognizing the necessity of considering multi-stakeholder awareness (e.g., government, citizens, construction departments, and the environment). These topics, therefore, align with the “institutions” dimension. Figure 13 shows the top-probability terms for each of the ten topics, along with the dimensions of a “Smart city” to which they belong.

• Topic Ratio Analysis.

Figure 14 is a composite pie chart showing the proportional distribution of the ten topics extracted across the three dimensions of a “Smart city”. The “technology” dimension ranks first with a total of 355 papers (58%) on six topics including “Topic 1”, “Topic 4”, “Topic 5”, “Topic 8”, “Topic 9”, and “Topic 10”. Next comes the “institutions” dimension with 161 papers (27%) on two topics (“Topic 2” and “Topic 3”), followed by the “people” dimension with 88 papers (15%) in two topics (“Topic 6” and “Topic 7”).

Figure 15 shows the percentage distribution of the ten topics extracted by the research stage. In Stage 1 (2007–2013), most of the research came from “Topic 2” (23.08%), which focused on building a “Smart city” from the perspective of urban planning and utilizing technologies, such as image processing, 3D visualization, and navigation algorithms. In Stage 2 (2014–2020), the highest percentage of topics shifted to “Topic 8” (18.64%), which focused on building basic frameworks and infrastructure using technologies such as “Augmented reality” and “Virtual reality”. In Stage 3 (2021–20 October 2023), the most popular research was “Topic 9” (19.73%), which involved applying the latest technology, “Digital twins”, to simulate urban architecture or models. In addition, it was observed that the topic related to the “technology” dimension consistently ranked first in terms of research volume regardless of the stage, followed by the “institutions” dimension and the “people” dimension. It is worth emphasizing that although the detailed topics related to the “technology” dimension were different depending on the stage, the research ratio continued to expand over time. In particular, “Topic 1”, “Topic 5”, and “Topic 9”, which include new popular terms such as “Digital twin” and “Metaverse”, showed a dramatic increase in research volume from Stage 1 to Stage 3. Among the remaining technology topics that include fundamental technologies (i.e., VR, communication, edge computing) for realizing the “Metaverse”, only “Topic 10” showed a significant increase in research volume, while “Topic 4” and “Topic 8” sharply declined.

5. Discussion and Conclusions

The “Metaverse” encompasses technologies related to virtualization and immersive digital environments and has attracted considerable attention from researchers and industry [75]. Despite the growing interest among researchers in the “Metaverse” within the context of smart cities, there remains a lack of comprehensive literature analyses using quantitative methods. Although a few studies have been conducted using bibliometric techniques, they are mainly descriptive and concentrate on technological elements. This trend limits a thorough understanding of the “Metaverse” and is particularly insufficient in providing a comprehensive view from various perspectives on how research on the “Metaverse” in the context of smart cities is progressing.

5.1. Answers to the Research Questions

To address RQ1, “What is the Metaverse and how is it being explored in the context of smart cities?”, this paper first conducted a literature review on the concept of the “Metaverse”, followed by a performance analysis based on the documents related to the “Metaverse” and “Smart city” retrieved from “WoS”. Regarding the definition of the “Metaverse”, this study agrees that the “Metaverse” is a type of technology related to VR [1,2,3]. While this technology has a 30-year history, it only began to truly blossom after Facebook’s rebranding to “Meta” in 2021 [7,16]. This development process shows that the implementation of the “Metaverse” is closely tied to technological development. In this context, as technology rapidly advances, research on the “Metaverse” has become more active and has made significant progress. Examining the results of our performance analysis, it seems that research on the “Metaverse” related to smart cities began to grow rapidly in 2014 and exploded after 2021. Additionally, the analysis results show that the most active countries in this research field are China, the United States, and Korea. The field with the most research is “Computer Science”, and the journal with the most papers is “Sustainability”. The researcher and institution who conducted the most studies are “Lv ZH” and the “Chinese Academy of Sciences”, respectively.

To answer RQ2, “Regarding the implementation of the “Metaverse” in the context of smart cities, what are the research trends, especially in the three dimensions of technology, people, and institutions?”, this paper performed science mapping consisting of keyword analysis and topic modeling. In the keyword analysis, this paper considered the similar development timeline of the “Metaverse” and “Smart city”, dividing the research area where these two topics intersect into three stages as follows: Stage 1 (2007–2013), Stage 2 (2014–2020), and Stage 3 (2021–20 October 2023). The keyword network mapping was conducted by focusing on the degree and betweenness centrality of keywords in each of these three stages. The analysis results in Stage 1 show that the “Metaverse” research related to smart cities mainly focused on “urban planning”. In Stage 2, the focus of the research shifted to a more integrated and holistic form of “urban governance”. In Stage 3, the research topics became more diverse and multidimensional, considering the perspectives of various stakeholders involved in building smart cities. Subsequently, through topic modeling analysis, this paper extracted ten topics based on the LDA algorithm and analyzed their relevance to the technology, people, and institutions dimensions. Among these extracted topics, six are classified into the “technology” dimension, two into the “people” dimension, and the remaining two into the “institutions” dimension. The analysis results of this topic modeling also show that in all three stages, the number of research topics in the “technology” dimension always ranked first, followed by the “institutions” dimension and then the “people” dimension. Additionally, this analysis also reveals that there was a significant increase in research on trending buzzwords such as “Digital twin” and “Metaverse”.

5.2. Future Research Directions on the Metaverse for Smart Cities

This study analyzed studies on the “Metaverse” in the context of a “Smart city” from various aspects using bibliometric techniques. By offering a comprehensive knowledge framework on the “Metaverse” research in smart cities, this paper provides valuable insights for academia to understand the current research landscape. Based on the research motivation and the bibliometric analysis results, this paper proposes the following three research directions to promote the flourishing of the field.

The first research direction is to balance research dimensions by focusing more on the “people” and “institutions” dimensions, which are relatively under-researched. Through keyword analysis and topic modeling, this study identified the existing studies mainly focused on the technical aspects of the “Metaverse”, while the dimensions of “people” and “institutions” are not sufficiently addressed. Previous studies have emphasized the importance of understanding a “Smart city” as an intelligent technological system made up of components from various domains with multi-stakeholders [128,129,130,131]. Based on this understanding, future research needs to move in a more balanced and comprehensive direction by paying more attention to the people and institutions dimensions. Especially when addressing the ongoing issue arising from both positive and negative environmental impacts brought by the “Metaverse” [101], namely, the conflicts among urban efficiency enhanced by the “Metaverse”, intensive energy and heavy resource demands, and the sustainability goals of a “Smart city”, future research should place more emphasis on the “people” and “institutions” dimensions. From the perspective of “people”, the study priority should move to how actively engaging citizens in decision-making processes and fostering a culture of innovation can lead to creative and inclusive solutions. By considering human involvement, technological progress can be balanced with responsible resource management, which helps address sustainability challenges and ultimately improves the quality of life for citizens [29,36,41,42,92]. Meanwhile, from the perspective of “institutions”, the research focus in the future should pay more attention to increasing the collaboration among multi-stakeholders (i.e., businesses, government agencies, and community organizations) [34,35] for creating a coordinated approach to managing resource demand while implementing the “Metaverse” in cities. By aligning the interests and efforts of these groups, long-term strategies for sustainable urban development can be facilitated, which eventually helps to mitigate the negative impacts brought by the “Metaverse”.

The second research direction is to address the concerns related to the “Metaverse”. Previous studies mainly focused on the benefits of the “Metaverse” in building smart cities while overlooking the associated risks. Implementing the “Metaverse” in the context of Smart city development requires integrating various underlying technology components (e.g., AI, IoT, blockchain, VR, AR, XR, MR). Typically, such technological integrations require the participation of multiple companies, which inevitably leads to compatibility and standardization issues among technical elements [2,7]. However, some technologies, such as IoT, inherently pose security and privacy risks [132,133]. As such, when implementing the “Metaverse”, it is necessary to pay close attention to security and privacy challenges. In addition, the “Metaverse” is also likely to form and promote complex social relationships that can raise ethical and moral issues in virtual societies while giving individuals new identities [2]. In addition, the “Metaverse”, the next-generation Internet [8], relies heavily on smart media and devices, which, if overused or mismanaged, can lead to digital addiction and mental health problems [62]. Therefore, future research should investigate privacy issues and mental health problems caused by the “Metaverse” to prevent its negative impact on users’ quality of life in smart cities. In particular, future research should prioritize investigating how to enhance security and privacy protection measures in the “Metaverse” to ensure a safe and inclusive virtual urban environment. Furthermore, research efforts should also attempt to standardize various technological components to guarantee seamless integration across industries in a Metaverse-driven city. Additionally, research on how to develop frameworks that foster the responsible and ethical use of the “Metaverse” will be another essential part of preventing misuse and ensuring accountability within the context of smart cities in the future.

The third research direction is to enhance the research methodology. This paper highlights a significant methodological gap in the existing studies [6,17,30,62], which rely heavily on theoretical descriptions and qualitative discussions with limited quantitative data analysis. In other words, the research volume related to using data-driven quantitative methodologies (e.g., bibliometric analysis) in the intersection field of the “Metaverse” and “Smart city” is quietly rare. Thus, future research needs to apply more advanced quantitative methodologies, such as advanced data mining techniques and AI technologies, to provide robust insights. Moreover, the hybrid research techniques that integrate both quantitative and qualitative methods should also be explored in future studies. By combining these approaches, researchers can gain a more comprehensive understanding of the intersection field of the “Metaverse” and “Smart city”, which aid in capturing both measurable data and deeper insights into user experiences, social interactions, and ethical considerations. In addition, the existing methodology lacks an interdisciplinary approach that integrates urban planning, sociology, governance, education, healthcare, and transportation. Therefore, employing such an interdisciplinary approach is essential in the future to address the challenges posed by the various aspects of the “Metaverse” in the development of a “Smart city”.

5.3. Research Contribution

Previous studies on the “Metaverse” in the context of smart cities, as listed in Table 1, have provided valuable insights on this topic. However, compared with these studies, this study has the following differentiating points, and these differences constitute the research contribution in this paper.

The first differentiating point is related to the research dimension. Kusuma and Supangkat [17], Bibri and Allam [6], and Yaqoob et al. [62] primarily focused on the solely “technology” dimension, emphasizing the utilization of enabling technologies (i.e., VR, AR, MR, XR, cloud computing, blockchain) to realize the “Metaverse” within the “Smart city” domain. On the other hand, Allam et al. [30] and Boulanger [61] started to pay more attention to sustainability and the perspectives of multiple stakeholders, which can be categorized under the “institutions” dimension. Compared with these previous studies, this research provides a more comprehensive and systematic analysis by integrating various dimensions—“technology”, “people”, and “institutions”—into a single unified framework, thereby offering a broader understanding of the implementation of the “Metaverse” in smart cities.

The second differentiating point is related to the research methodology. Although most previous representative studies have relied on literature-based research, they predominantly utilized qualitative techniques such as literature reviews and systematic reviews rather than scientific quantitative methods. Among five representative studies in Table 1, only Boulanger [61] employed a quantitative approach through bibliometric analysis, but it was limited to basic mapping techniques. In contrast, this study systematically applies bibliometric techniques, including performance analysis and science mapping, thereby conducting a more comprehensive and scientific literature study.

5.4. Limitations of the Study and Future Research Directions

The limitations of this study are as follows. First, the search time and source scope of the data used in this study can be noted as limitations. Specifically, regarding the search time, the period of literature collection from “WoS” in this paper was limited to 2007 to 20 October 2023. As for the source scope, this paper collected data only from the “WoS” database. Although the “WoS” database is globally renowned for its large number of quality-assured publications, this paper did not utilize materials from other databases such as “Google Scholar” and “Scopus” as sources, which may lead to retrieval omissions. Second, there is a limitation regarding the search strings used. This study employed various search strings in the “WoS” database, as specified in Table 4. However, it is challenging to cover all relevant terms at once, given the large number of synonym substitutions that can be traced in the literature. Moreover, with rapid technological developments and the continuous emergence of new terms driven by changing policies and environments, the search strings used in this paper may not sufficiently capture related technical terms in the future. Third, another limitation of this paper arises from the quantitative methodology of bibliometric analysis itself. While this approach emphasizes quantitative metrics such as publication counts, keyword networks, and topic modeling analysis, which are crucial for mapping research performance and trends across the three dimensions (technology, people, and institutions), they do not capture the qualitative advancements in these fields.

To address these limitations, this paper also suggests future research directions. First, future research should consider incorporating multiple databases, such as “Google Scholar” and “Scopus”, and extending the retrieval period to ensure continuous updates and improvements to the research findings. Second, future studies should focus on updating and refining the search strings to reflect the latest developments accurately. Third, future research should not neglect the introduction of qualitative methods while expanding the employment of quantitative techniques to assess the deeper advancements in the three dimensions (technology, people, and institutions) in smart cities. By integrating both quantitative and qualitative methods, future studies can provide a more comprehensive and in-depth understanding of progress in the intersection area of the “Metaverse” and “Smart city”.