Emerging Trends in Interactive Space: A Scientometric Analysis

Abstract

With the advent of the Fourth Industrial Revolution and the rise of new forms of productive forces, the ways humans interact with space, objects, and information are being profoundly reshaped, bringing unprecedented possibilities for upgrading interactive spaces—human settlements that integrate physical and digital environments. Against this background, using the literature on interactive space research from the Web of Science (WoS) Core Collection between 1990 and 2025 as the data source, this study employs CiteSpace software to generate scientific knowledge maps, analyzing the historic development, hotspots, and trends in the research of interactive space, providing both theoretical and data support. In terms of results, a total of 458 papers were collected, demonstrating a consistent year-on-year increase. The research spans multiple fields, including computer science, architecture, ecology, physics, design, and behavioristics. Specifically, results indicate that research hotspots in interactive spaces include collaborative governance, social coexistence, and sustainable renewal, all of which are highly relevant to activating human settlements. The vitality of interactive spaces can be constructed across multiple dimensions, (for instance, enhancement based on ecology, environment, culture, and other factors of the space). However, research on interactive spaces still suffers from a lack of interdisciplinary collaboration and multi-domain integration; therefore, it is essential to strengthen cooperation among relevant fields. Current research lacks interdisciplinary integration and dynamic response mechanisms. Based on these findings, this study, through visual analysis, reveals the research hotspots and evolutionary trajectory of interactive spaces and proposes a “technology–humanism–governance” trinity framework. This system should be based on technology as the means, humanism as the guiding principle, and effective governance as the goal. It aims to explore how to leverage the service-oriented and convenient nature of technology in interactive spaces to deepen human-centric design and thereby drive the optimization of systems. Based on these findings, future research on interactive spaces should shift its design philosophy to be more human-centric, establish a multidisciplinary research system, utilize local empirical cases, and develop scalable, applicable theories to construct harmonious, open spaces, enhance human–environment relationships, and provide other countries undergoing urbanization with practical solutions.

1. Introduction

An interactive space is a dynamic response system that utilises digital technology as its neural centre, the physical environment as its carrier, and human behaviour as its driving force [1,2]. Since Google first introduced the concept of cloud computing in 2006, the accelerated integration and application of big data, artificial intelligence, blockchain, and other technologies have equipped urban governance with perception, cognition, and decision-making capabilities, facilitating the transition of cities from “managed physical spaces” to “digitally integrated urban environments.” Consequently, interactive spaces have evolved from basic “information displays” to “immersive intelligent agents” [3]. In the current research on sustainable urban development and public spaces, although the concept of “interactive space” has received much attention, its connotation, evolution, and frontier issues still lack systematic clarification. To bridge this cognitive gap, this study employs the CiteSpace bibliometric method to conduct a knowledge graph analysis of related domestic and foreign research, revealing that the achievements in this field are concentrated in technical dimensions such as spatial form design, while there is a significant absence of the promotion of social inclusiveness and humanistic care as the core proposition, especially regarding human-oriented sustainable mechanisms. This finding provides empirical evidence and a key entry point for clarifying the research logic starting point and promoting the theoretical shift from “humanistic deficiency” to “interactive symbiosis”.

These findings have prompted us to conduct a thorough examination of current digital practices and to reflect on the numerous issues involved in these practices when striving for truly interactive urban spaces. At the 2025 Global Digital Economy Conference, the core agenda focused on “Building International Cooperation Platforms and Promoting the Integration of Digital Technology and Urban Space,” aiming to empower smart city construction through digital technology. Internationally, Singapore’s “Virtual Singapore” project has established a nationwide digital twin interactive system [4]; the EU’s “Living Lab” initiative promotes citizen participatory space design [5]; and China’s “City Brain” project focuses on government–citizen interaction scenarios [6]. Digital empowerment through next-generation information technologies reconstructing urban operational mechanisms provides greater potential for enhancing governance efficiency and public service levels. The interactivity of space is gradually shifting towards digital empowerment; that is, through technological transformation, it enhances the human-centric quality of space to achieve its interactivity, thereby promoting the sustainable development of space. A deeper ethical and cultural reflection is thus imperative, not only to scrutinise how technology choices shape social interactions and power dynamics but also to deliberate on what kinds of technologies should be adopted and how they can be aligned with principles of justice, transparency, and cultural diversity. Issues such as the lack of human-centricity, resource misallocation, and environmental pollution during urbanisation need to be addressed and resolved. By reconstructing an “interactive symbiosis” model, it offers theoretical paradigms and practical pathways for upgrading urban spaces to be affordable, inclusive, and highly engaging, responding to the urgent demands of UN SDG11 for sustainable urban development [7].

The necessity of this study is rooted in two critical unresolved gaps: The first is the lack of dynamic responsiveness. Existing models rely on predefined scenarios (e.g., Barcelona’s Superblocks [8]) and cannot adapt to sudden public events (such as spatial and functional transitions during a pandemic [9]). The second is an imbalance in empowerment: corporate-dominated systems (e.g., Amazon Sidewalk [10]) reduce citizens to data providers rather than decision-making participants. In the current social context, we propose a spatiotemporally coupled digital twin interaction architecture; establishing a comprehensive research system is crucial for sustainable human development. Therefore, this paper utilises CiteSpace to analyse the research trends and knowledge graph structure in this field. This paper proposes the concept of future optimisation, a unique research direction, and further expansion of the overall framework, as well as the basis and suggestions for the study of interaction space, all of which provide guidance and practical experience for subsequent research.

The structure of this paper is as follows. The first part introduces the data sources and research methods. The second part analyses relevant public space policies and provides case studies of interactive spaces. Then, a review is conducted based on the number of published papers, author collaborations, publishing journals, and institutional collaboration distributions. Subsequently, for the most influential categories within the research field, highly cited articles and literature are listed for analysis and summary, and the latest research hotspots and future research directions are identified through keyword analysis. Finally, this paper discusses future research hotspots, evolutionary trajectories, and prospective developments in the field.

Collectively, the necessity of this research stems from two critical gaps: the deficiency in dynamic response capabilities and the imbalance in improper allocation of rights and responsibilities. To address these, we propose a spatiotemporally coupled digital twin interaction architecture and conduct a bibliometric analysis utilizing CiteSpace, through which research hotspots, evolutionary trends, and future directions are systematically organized to provide guidance for subsequent investigations.

2. Literature Review

2.1. Definition of Interactive Space

An interactive space is an intelligently responsive system that uses digital technology as its neural framework, the physical environment as its structural medium, and human activity as its primary driver. Moving beyond the conventional perception of space as a static or passive ‘container,’ it uses technologies such as the Internet of Things (IoT), artificial intelligence, real-time sensing, and data analytics to transform built environments into adaptive organisms that can perceive, process, and respond to dynamically changing human needs. Its fundamental innovation lies in transcending the traditional “container-based” paradigm in architectural and spatial design, which often rigidly couples form and function. Instead, it emphasises plasticity, interactivity, and experiential quality. Central to this redefinition is departing from classical spatial cognition toward a new ‘space-as-a-service’ paradigm. As the author of the reference explicitly puts it, ‘This work aims to contribute to the concept of public space as a service,’ which emphasises more convenient and human-centric technological services, reflecting a new type of spatial interactivity [11]. In this model, space is an on-demand, customizable, and upgradable service interface, not merely a physical locale. It adapts in real time to behaviors, emotional states, and intentions of occupants—modulating lighting, thermal conditions, spatial configurations, and even informational content—to achieve functional efficiency and affective resonance between spatial resources and human activities. It consequently reshapes the relationship between people, technology, and the environment, guiding architectural and spatial design toward intelligent, human-centric, and responsive trajectories [12,13,14].

2.2. Development Trends and Changes in Interactive Spaces

The conceptual evolution of interactive spaces reflects a trajectory from techno-centric infrastructure to human-centric experiential environments, increasingly characterised by fluid physical–digital integration and intelligent adaptability [15,16,17]. This article categorizes this evolution into three developmental phases, each marked by distinct technological breakthroughs, theoretical paradigms, and representative projects.

(1) The Enlightenment Period (1990s–2000s) [18]: stemming from Weiser’s vision of ‘ubiquitous computing,’ space began to be conceptualised as ‘silent infrastructure.’ (2) The Awakening Period (2010–2015) [19]: this phase was marked by Greenfield’s notion of “programmable space” in Smart Cities and experimental interactive projection systems, such as those developed at the MIT Media Lab. (3) The Fission Period (2016–present) [20] is characterized by the emergence of the metaverse, which facilitates a deeper fusion between physical and virtual spatial entities.

Although these developmental stages have demonstrated the potential of interactive spaces, significant methodological and conceptual gaps remain. Current research lacks integrated, multi-stakeholder frameworks and struggles to address core challenges, such as sensing dynamic demand and reconfiguring spaces adaptively [21,22,23]. Consequently, many existing systems exhibit “static intelligence” traits, failing to capture the richness of human experience and adapt to real-world complexities [24,25,26]. This gap underscores the urgent need for more holistic, adaptive, and ethically grounded approaches—a challenge that this article aims to address.

2.3. Current State

Currently, the interactive space is experiencing active technological exploration but fragmented application development on a global scale. However, it still faces significant bottlenecks in dynamic adaptability, integration with social systems, and cross-disciplinary collaboration. For example, projects from the MIT Senseable City Lab excel in real-time data visualization but often overlook underlying social and behavioral mechanisms [27]. Research trends indicate that the field is gradually shifting towards collaborative governance and sustainability. However, theoretical progress lags behind in practice, and scalable solutions that prioritize people have yet to be systematically developed [28].

2.4. Gaps in the Literature

Despite advances, critical research gaps remain. Current studies lack holistic frameworks integrating technological, social, and experiential dimensions, often prioritising technical implementation without social context [29] or spatial theory without actionable digital strategies [30]. Moreover, while real-time responsiveness is emphasized, genuine dynamic adaptation remains rare due to limitations in sensing, algorithmic, and interpretive capabilities [31]. Many systems rely on pre-set interaction templates [32,33], resulting in performative adaptation rather than deep responsiveness.

2.5. Innovative Features

This study is distinguished by several key innovations. The first is a multi-scalar methodology combining immediate adjustments with long-term spatial learning based on cumulative occupancy and cultural patterns. This methodology enables environments to evolve beyond short-term reactivity. Second, the study embeds ethical foresight and participatory design into the core of its development process. This promotes “justice-aware interfaces” that operationalize equity as a technical requirement. Thus, this study bridges the gap between theoretical critique and practical strategies.

2.6. Future Research Directions

Future research on interactive spaces should aim to create intelligent, inclusive, and resilient environments. This includes developing multimodal sensing to interpret subtle behaviors [34,35,36], advancing explainable and adaptive AI for anticipatory learning [37], and integrating interdisciplinary approaches from behavioral science and urban design [38,39]. New metrics are needed to evaluate social and ethical outcomes such as spatial fairness [40]. Additionally, adopting decentralized, blockchain- or edge-based governance will enhance user agency and co-creation as interactive spaces merge with immersive platforms [41].

3. Data Source and Research Methods

3.1. Data Source

This study utilises the Web of Science platform as its data source. In 1997, Thomson Reuters integrated the Science Citation Index (SCI), Social Sciences Citation Index (SSCI), and Arts and Humanities Citation Index (AHCI) to create an online version of its interdisciplinary literature database—Web of Science—leveraging the open Internet environment. This study focuses on interdisciplinary theoretical innovation, requiring the literature sample to simultaneously cover three major fields: computer science (human–computer interaction), urban planning (smart cities), and sociology (spatial governance). As a global database, WoS offers comprehensive coverage across all fields and compatibility with bibliometric analysis. Its authority and scientific rigour establish it as an objective and reliable research foundation.

A specialised search method was employed, using an exact phrase query TS = (“interactive space” OR “interactive spaces”) restricted to subject fields (including titles, abstracts, and keywords) rather than full-text search. This approach ensures research relevance while filtering out extraneous content [42,43]. All literature data used in this study were retrieved from the Web of Science (WoS) Core Collection database, covering publications from 1 January 1990 to 17 July 2025. To achieve the most comprehensive coverage of the field, no additional restrictions were imposed on document types. This inclusive strategy allowed the retrieval of the broadest possible set of relevant studies. All 458 relevant original articles were retained and exported in plain text format as “full records and cited references.” As the search was strictly limited to the subject term “interactive space(s),” usage habits and preferences for this term across different countries, institutions, and disciplines may introduce bias into the literature sample. For instance, certain regions or research fields may favour this specific phrasing, while other relevant studies might employ different terminology. Consequently, comprehensive coverage is inevitably constrained.

3.2. Research Methods

CiteSpace is a scientific bibliometric analysis tool used for literature reviews [44]. It is a Java-based information visualization and knowledge mapping software that presents the structure, patterns, and distribution of scientific knowledge through visual analysis and developed by Chaomei Chen [45]. It is used worldwide for identifying emerging topics and frontier trends in hotspot research. The generated charts are called “knowledge maps” or “bibliometric maps.” We analyzed 458 literature entries exported from WoS in “full-text records and cited references” format. Duplicate records were removed using the built-in deduplication function in CiteSpace (version 6.2.R6) during data import. The function automatically detects and excludes identical entries. To reduce noise in collaboration and co-citation networks, author and institution names were standardized by combining CiteSpace’s built-in name-merging function with manual correction. After screening and sorting, literature types such as reviews, book chapters, and news articles were removed, leaving 450 valid documents for analysis [46]. In the current study, time slicing was set to one year, and the other settings were left at their defaults. This “default setting” approach is a standardized configuration designed to filter out noise and reveal the key paths [47]. Key information—such as keywords, authors, countries of publication, research institutions, and publishing journals—was selected and analyzed. Through analyzing the relationships between various thematic elements, this study clarifies the current development status and evolutionary trajectory from 1990 to the present and provides a data-driven forecast of future research directions.

3.2.1. Bibliometric Tools and Settings

This study used CiteSpace (version 6.2.R6) and Excel to improve the multidimensionality and precision of the analysis. CiteSpace (version 6.2.R6) was used to construct and cluster the literature network, and the CiteSpace parameter settings are as follows [48]:

(1)

The time period was set to 1990–2025 (the full period), with one year per slice and selecting 50 high-frequency nodes (Top N = 50) per slice.

(2)

The k-value was flexibly adjusted to suit the requirements of different analysis types in order to optimise the clarity and structural presentation of the graph based on the g-index algorithm.

(3)

The ‘Pathfinder’ and ‘Pruning the merged network’ settings were used for network structure pruning.

(4)

The log-likelihood ratio (LLR) clustering algorithm was used for cluster analysis to extract noun terms from the titles of the documents and name the corresponding research contents [49].

3.2.2. Visualisation and Network Construction

This research explores interactive spaces at the intersection of digital technology and sustainable development. As shown in Figure 1, the images generated by CiteSpace demonstrate a clustering modularity of Q = 0.8648 (>0.3 indicates significance) and a silhouette of S = 0.9667 (>0.7 indicates high confidence); values above 0.5 suggest reasonable clustering; and values above 0.7 indicate highly reliable results, thereby confirming that the sample largely represents the multidimensional nature of the research subject. Before the literature screening process, the time window was set from 1990 to 2025, covering the complete technological evolution cycle of interactive spaces. The CiteSpace network diagrams were supplemented with an Excel-based analysis of publication trends and geographic distribution.

3.3. Research Design Process

We searched for “Interactive Space” and “Interactive Spaces” on 17 July 2025. In the WoS core collection, TS = “Interactive Space” OR “Interactive Spaces.” These records were exported in “Plain Text” and “Full Record with Cited References” formats. They were then imported into CiteSpace (version 6.2.R6) software for analysis. This paper’s relationship diagram shows the different fields obtained through CiteSpace from 1990 to 2025. In the current study, “interactive space” was used as the keyword and “evaluating ‘interactive space’” as the search term. Time slicing was set to one year, and the other settings were left at their defaults. Key information—such as keywords, authors, countries of publication, research institutions, and publishing journals—was selected and analysed. Through thematic relationship analysis, this study clarifies the current status, traces the evolutionary trajectory (1990–present), and offers a data-driven forecast of future research directions in interactive spaces. Figure 2 presents the research process design outline.

As with most bibliometric studies, this research has inherent limitations. Relying solely on the Web of Science Core Collection (WoSCC) introduces a language bias by prioritising English-language journals. This approach may potentially exclude regional or non-English publications and grey literature, such as reports and policy documents [50,51]. Furthermore, differences or overlaps between WoSCC and other databases (e.g., Scopus and PubMed) may affect coverage. These inherent limitations are explicitly stated and further discussed in Chapter 5 to ensure transparency.

4. Results

In recent years, the United Nations system has been continuously promoting the empowerment of global sustainable development through digital technology, forming a strategic framework with clear objectives and multi-level coordination [52,53,54,55]. This series of policy documents and initiatives primarily advocates for artificial intelligence, virtual reality, and smart city technologies as key drivers in building a safe, trustworthy, and inclusive digital future [56,57,58,59,60]. As shown in

Table 1. Interactive space-related policy documents and implementation standards.

Document	Release Time	Issuing Authority
The 10th Edition of the “Sustainable Development Goals Report”	July 2025	United Nations
“The Global Action Plan on Artificial Intelligence Governance”	July 2025	United Nations
“The Paris Agreement”	November 2024	United Nations Framework Convention on Climate Change
“The Resolution on Strengthening International Cooperation in Building Capabilities of Artificial Intelligence”	July 2024	United Nations
“Seizing the Opportunities Brought by Safe, Reliable and Trustworthy Artificial Intelligence Systems and Promoting Sustainable Development”	March 2024	United Nations
“International Guidelines for Smart Cities Based on a People-Centric Approach”	June 2023	United Nations Human Settlements Programme, WeGo and UNECE
“Report on Sustainable Smart Cities”	December 2020	ITU,
Global Initiative of Sustainable Smart Cities (U4SSC)	Started in 2019	United Nations
“The 2030 Agenda for Sustainable Development”	September 2015	United Nations
“Global Governance Action Plan for Artificial Intelligence”	July 2025	Related UN agencies
“Global Digital Inclusive Initiative (Draft)”	July 2025	18 UN entities
“Call to Action for United Nations Virtual World Day”	June 2025	United Nations Educational, Scientific and Cultural Organisation, UNESCO
“National Information Plan 2023–2029 Strategic Plan”	Started in 2023	United Nations Development Programme, UNDP
“Digital Strategy 2022–2025”	January 2023	United Nations Creative Department
UNVR Virtual Reality Applications	August 2025 updated	United Nations Institute for Training and Research
UNITAR Metaverse Initiative	November 2023 started	United Nations

, the 2030 Agenda for Sustainable Development and its subsequent assessments (e.g., the 10th progress report in 2025) established the fundamental goals of sustainable development [61,62,63,64,65,66]. Building on this, the Global Action Plan on AI Governance (2025) collectively establishes the core principles of global AI governance [67,68,69,70]. It emphasizes that AI development must serve the Sustainable Development Goals (SDGs) and ensure safety, reliability, and trustworthiness [71,72,73]. Simultaneously, the “People-Oriented Smart Cities International Guidelines” (2023) and reports from agencies such as U4SSC and UNECE offer operational guidelines for translating ethical principles into specific planning and construction standards at the city level [74,75]. The Global Digital Friendly Initiative (draft, 2025) and the UN Virtual World Day Call to Action (2025) further call for the creation of an equal, safe, and friendly digital environment for all, especially youth [76,77,78]. We cite these policy documents in order to demonstrate that the core topic of this article has indeed received corresponding attention at the highest level of the policy agenda, thereby verifying the forward-looking quality and significance of the research topic and findings of this study.

At the level of technology application and project demonstration, various UN entities are actively utilising immersive technologies to enhance their public service and communication effectiveness [79,80]. Pioneering projects like UNVR’s virtual reality applications (2025 update) [81] transcend conceptual discussions by directly demonstrating the significant potential and practical value of technologies such as interactive spaces and virtual environments in areas like crisis response, public education, skills training, and cultural heritage preservation [82,83,84]. These projects provide member states with real-world examples, proving how responsible innovation can contribute to achieving the SDGs [85,86,87]. Together, these policies and projects convey a clear message: Future development must be technology-empowered, human-centric, and globally collaborative [88,89,90,91,92]. Digital technology, especially AI and interactive spaces, is now a core infrastructure and an innovative mean of supplying global public goods and accelerating the achievement of the SDGs [93,94,95].

4.1. Overview of Research Progress

4.1.1. Publication Volume (Temporal Distribution)

Publication volume, as a core indicator in bibliometrics, holds statistical significance far beyond mere quantity accumulation. This analysis aims to quantify the academic output in the field of interactive space from 1990 to 2025, as shown in Figure 3, and achieve three goals. The first is to reveal the discipline’s life cycle by identifying key inflection points from technological germination to theoretical explosion. The second is to map technological–societal coevolution by correlating publication peaks with major technological breakthroughs, policy shifts, and national verification strategies. The third is to predict research frontiers by capturing the burst threshold of emerging topics through changes in growth rates.

Based on the publication volume chart shown in Figure 4, the core findings can be divided into a three-stage evolutionary trend, detailed as follows.

1990–2004: The Growth Period, characterised by the challenging development of technological idealism. This period reflects a certain degree of disconnection from practice due to conceptual advancement. The literature was concentrated in the computer field, highlighting disciplinary singularity.

2005–2013: The Key Development Period, marked by a surge in publications in 2005. To a large extent, as the number of global mobile devices skyrocketed from millions to 7 billion, the Internet, social media, and mobile devices became deeply integrated at both the technical and practical levels, seamlessly “weaving” into our daily lives and thereby redefining the “interaction space” [96]. This increase was strongly correlated with mobile Internet penetration rates and the implementation of 127 global smart city projects, establishing a paradigm within the smart city movement [97,98,99,100].

2014–2025: The Rapid Development Period, as evidenced by the ACM Digital Library database, shows that the metaverse interaction literature grew by 480% [101], and the proportion of highly cited literature exceeded 50% from 2021–2023. As shown in Figure 5, the overall number of published documents shows an upward trend.

To present the dynamic evolution of academic attention in this field more clearly, Figure 5 focuses on the period from 2001 to 2025 when the number of publications began to show a sustained increase, and the dashed line in the figure clearly indicates this trend. This trajectory highlights the growing academic significance of the field and its increasing relevance in the context of the challenges posed by global cross-border interactions.

4.1.2. Distribution of Publishing Journals

Journal distribution statistics provide a precise mapping of the knowledge production landscape of a discipline [102]. As shown in Figure 6, this study aims to reveal the dominant fields and discourse power distribution in the field and detect the depth of disciplinary integration through the proportion of interdisciplinary journals by analysing the distribution carriers of the 450 documents.

The co-citation network was assessed through node frequency, link strength, and density to generate a ranking of influential journals [103].

As shown in

Table 2. Journals with a high number of publications related to interactive spaces from 1990 to 2025.

Number of Published Papers	Release Time	Journal Title
64	1999	LECT NOTES COMPUT SC
31	1998	COMMUN ACM
20	2020	SUSTAINABILITY-BASEL
19	2014	INT J HUM-COMPUT ST
19	2019	PLOS ONE
17	2007	PERS UBIQUIT COMPUT
16	2020	IEEE ACCESS
16	2006	ACM T COMPUT-HUM INT
15	2017	34TH ANNUAL CHI CONFERENCE ON HUMAN FACTORS IN COMPUTING SYSTEMS
15	2001	P NATL ACAD SCI USA

, the top five journals with the most publications are Lecture Notes in Computer Science (the computer science series and conference proceedings published by Springer), Communications of the ACM (the flagship monthly publication of the Association for Computing Machinery in the United States), Sustainability (an open access journal published by MDPI, headquartered in Basel, Switzerland), the International Journal of Human–Computer Studies (one of the top journals in the field of human–computer interaction, published by Elsevier), and PLOS ONE (an interdisciplinary open access journal published by the Public Library of Science), with a total of over 232 articles published.

In the CiteSpace analysis, as shown in Figure 7, the node type was set to “cited journal,” and the time slice was set to one year to generate a map of co-cited journals. There were 777 nodes in the journal distribution, 2457 links, and a density of 0.0081. The more links there are, the more frequently the two journals appear together; The deeper the connection between them, the more the line colour shifts from blue to orange; the thicker the connecting line, the stronger the relationship between them.

Publications in the field of interactive spatial research exhibit pronounced disciplinary biases. The Computer Science Lecture Notes (Springer Publishing) command a substantial share, followed by high-impact journals such as IEEE Access, Nature, and Scientific American. As shown in Table 2, this distribution underscores computer science’s dominance, where research primarily focuses on the technical implementation of interactive spaces—creating and optimizing physical–digital convergence experiences through sensors, data processing, and interactive interfaces. In contrast, topics such as public governance and cultural adaptation within spatial contexts remain understudied, indicating a significant interdisciplinary research gap.

4.1.3. Distribution of Regional Collaboration

The analysis of regional collaboration distribution aims to deconstruct the power structure of global knowledge production, revealing resource allocation and collaboration patterns in the field of interactive space research. It is also noteworthy that a significant proportion of older records in the WoSCC are missing complete author address information, as documented in bibliometric methodology studies [104]. This absence may partly explain the sparse representation of certain countries in earlier decades and should be considered when interpreting long-term collaboration patterns.

By quantifying the strength of collaboration networks between countries and regions, the analysis locates the countries that dominate international cooperation and reveals the direction of resource flow. From 1990 to 2025, as shown in Figure 8, the top publishing region by citation count is the USA, with 78 citations. Second is PEOPLE’S R CHINA, with 69 citations. Third is SPAIN, with 35 citations. Fourth is ENGLAND, with 33 citations. Fifth is GERMANY, with 28 citations.

As shown in Figure 9, CiteSpace was used to analyse the distribution of collaboration in interactive space research. The regional collaboration distribution map contains 64 links, 127 nodes, and a density of 0.063. In CiteSpace visualisations, countries with high centrality are denoted by purple circles, indicating their greater research impact. The circle diameter corresponds to the count value. Betweenness centrality was used to evaluate the influence of a certain country, with higher values representing stronger bridging roles in international collaboration [105].

As shown in Figure 10, the timeline map chronologically organises bibliometric data—including keywords, subject terms, and citations—to visually trace the evolution of international collaborative research in interactive spaces. It highlights a pivotal shift between 1996 and 2006. Node proximity indicates stronger cooperation, while larger nodes reflect greater influence. This visualisation helps track evolving research frontiers, core theme transitions, and cross-national knowledge flows.

Overall, the USA leads in theoretical innovation and high-impact output, serving as a central hub in global research networks. Subsequently, PEOPLE’S R CHINA has also demonstrated strong scientific research capabilities and continuous investment in this field.

4.1.4. Distribution of Author Collaboration

Based on the researcher collaboration network map generated by CiteSpace, this study conducted a quantitative analysis of core scholars and their academic influence in the field of interactive space. As shown in Figure 11, the map presents the academic ecology through three dimensions: node size represents author citation volume, node colour distinguishes the research cluster (reflecting sub-field direction), and lines between nodes reveal collaboration relationships. For example, Ye, Qiongwei and Cerezo, and Eva belong to the emerging technology application cluster, while Eng, K and Verschure, and PFMJ belong to the early human–computer interaction theory cluster. This suggests generational differences in research themes. Imbalances in collaboration distribution reveal the rapid rise of emerging scholars and the increase in highly cited literature over the past three years. This could imply that the field is currently experiencing a phase of rapid theoretical development. Regional clustering of core author groups suggests the need for a more open academic community to promote knowledge flow.

As shown in Figure 12, co-citation frequency reveals the pivotal position of literature in the knowledge network. Highly co-cited literature constitutes the “intellectual bedrock” of the field’s development. The co-citation network map constructed by CiteSpace shows that research on interactive space has significant generational characteristics. The top five documents by citation frequency are the foundational research by WEISER M (15 citations), the technical framework by JOHANSON B (9 citations), the user behavior model by VENKATESH V (6 citations), the methodological innovation by CRESWELL JW (6 citations), and the interaction prototype by BALLAGAS (5 citations). The fact that Weiser’s seminal theory on ubiquitous computing has been co-cited across 15 subsequent studies suggests that its principles remain broadly relevant and continue to provide critical theoretical support for a wide spectrum of multi-branch research endeavours.

4.1.5. Distribution of Institutional Collaboration

By analysing the collaboration distribution between research institutions, as shown in Figure 13, an in-depth understanding of the field’s recognition of academic support was gained, thereby promoting collaboration between institutions. This approach facilitates the analysis of the influence of the institutions in this field to promote the development of the research field [106]. Collaboration between various research institutions promotes in-depth research in the field. The image generated by CiteSpace shows 260 nodes and 160 links, with a density of 0.0048. Node size is proportional to citation volume, and line thickness represents collaboration strength.

As shown in

Table 3. Top 10 in terms of citation frequency in institutional collaboration clustering.

Citation Counts	Institutional Collaboration
17	Yunnan University of Finance & Economics, 2017
10	Yunnan University, 2017
9	Aarhus University, 2006
7	Swiss Federal Institutes of Technology Domain, 2003
7	Aalborg University, 2005
6	Communauté Université Grenoble-Alpes, 2005
6	Universite Grenoble Alpes (UGA), 2005
6	Centre National de la Recherche Scientifique (CNRS), 2006
6	ETH Zurich, 2003
6	Massachusetts Institute of Technology (MIT), 1998

, the top five institutions by citation volume are Yunnan University of Finance and Economics, 2017 (17 citations); Yunnan University, 2017 (10 citations); Aarhus University, 2006 (9 citations); Swiss Federal Institute of Technology, 2003 (7 citations); and Aalborg University, 2005 (7 citations). These universities contribute substantially to publication output and play bridging roles in connecting different research communities.

As shown in

Table 4. Top 10 institutions in institutional cooperation clustering.

Degree	Institutional Collaboration
13	Communaute Universite Grenoble Alpes, 2005
13	Universite Grenoble Alpes (UGA), 2005
12	Centre National de la Recherche Scientifique (CNRS), 2006
10	Institut National Polytechnique de Grenoble, 2005
10	Universite Federale Toulouse Midi-Pyrenees (ComUE), 2016
8	Yunnan University of Finance & Economics, 2017
8	Technische Universitat Dresden, 2012
7	Swiss Federal Institutes of Technology Domain, 2003
7	University System of Georgia, 2002
7	CNRS—Institute for Information Sciences & Technologies (INS2I), 2016

, the top six institutions by collaboration strength (degree centrality likely) are Communauté Université Grenoble Alpes (2005, degree 13), Université Grenoble Alpes (2005, degree 13), CNRS (2006, degree 12), Grenoble Institute of Technology (2005, degree 10), Université Fédérale Toulouse Midi-Pyrénées (2016, degree 10), and Yunnan University of Finance and Economics (2017, degree 8). It is noteworthy that Communauté Université Grenoble Alpes has collaboration strength (13), thereby confirming its role as a pivotal connector across clusters. Distinguishing cluster colours shows two major research camps: Chinese institutions are concentrated in the emerging technology cluster, while European institutions are distributed in the basic theory cluster. This leads to gaps in cross-field collaboration to a certain extent. This also indicates that interactive space research is in a period of global resource reorganisation and urgently needs to establish a three-dimensional collaboration system across generations, regions, and fields.

4.2. Research Impact and Field Distribution

Analysing the academic significance of research fields allows us to parse the potential value of interactive space in theoretical exposition and field status. In this research, we analyse the interdisciplinary integration map, disciplinary betweenness, and thematic cluster evolution of interactive space in spatiotemporal dimensions for interactive space.

4.2.1. Highly Cited Articles

Co-citation analysis based on CiteSpace reveals the knowledge structure of interactive space research. Dual metrics of citation volume and centrality aid in identifying highly influential and structurally pivotal literature. As shown in Figure 14, network visualisation presents distinct patterns; Johanson B (2002) [107] anchors the core knowledge framework, connecting other key developments through Ballagas R (2003) [108] and Bernardet U (2002) [109]. Davis JW (1997) [110] similarly serves as a bridge.

As shown in

Table 5. The top ten cited references are ranked by their node degrees.

Degree	Release Time	Cited Articles
26	2003	Ballagas R, 2003, CHI, V0, P537 [108]
21	1996	ABOWD G, 1996, P ACM MULT 96 C, V0, P0 [111]
21	1996	Bobick A, 1996, KIDSROOM PERCEPTUALLY, V0, P0 [112]
20	1996	BRAY T, 1996, 5 INT WORLD WID WEB, V0, P0 [113]
20	1999	CHEN C, 1999, INFORMATION VISUALIS, V0, P0 [114]
18	2002	Bernardet U, 2002, NEUROCOMPUTING, V44, P1043 [109]
16	2003	Hinckley K, 2003, PROC. UITS03, V0, PP149 [115]
16	2000	Brumitt B, 2000, LECT NOTES COMPUT SC, V1927, P12 [116]
16	2003	Bailey B, 2003, P ACM C HUMAN FACTOR, V5, P313 [117]
15	1989	AVATARE A, 1989, SICST898917 SWED I C, V0, P0 [118]

, three seminal early works shaped interactive space research: ABOWD G (1996) [111] established user-centred HCI principles; Bobick A (1996) [112] advanced AR for digital–physical integration; and Johanson B (2002) [107] synthesised these into a unified framework. Together, they converged computer vision, graphics, HCI, and AR, defining the field as an interdisciplinary study of seamless human–environment interaction.

The core literature primarily concentrates on the period between 1996 and 2003, marking the peak of knowledge integration during this era. A significant gap emerged after 2003, characterised by an absence of highly influential publications. The notable absence of highly cited literature after 2003 may stem from the following factors.

(1)

Evolution of terminology: initial terms such as “ubiquitous computing” or “perceptive environments” gradually shifted toward “smart environments,” “IoT,” or “embodied interaction,” potentially dispersing citations across different literary clusters.

(2)

Maturation and specialisation: as the field grew, research branched into subdomains (e.g., tangible UI, ambient intelligence), reducing the dominance of earlier umbrella terms.

(3)

Database bias: older seminal works may continue to be cited as canonical references, while mid-2000s publications might belong to more specialised veins that are highly cited within sub-communities but not across the entire field.

(4)

Technology readiness: widespread adoption of sensing technologies, mobile devices, and IoT infrastructure did not occur until the 2010s, which may have delayed large-scale applied research that would yield new high-impact papers.

4.2.2. Research Fields

As illustrated in the discipline co-occurrence network time zone map produced by CiteSpace (Figure 15), the visualisation conveys two key metrics through its node design. Firstly, the node size corresponds to the frequency of occurrence. Secondly, the thickness of the purple outer ring serves as a proxy for a node’s influence; a thicker border indicates a greater academic impact within the network structure.

This study employs field centrality metrics to identify knowledge hubs within the field of interactive space research. Key domains with high centrality scores (>0.15)—including electrical engineering (0.19), cybernetics (0.17), and communication science (0.16)—serve as critical knowledge conduits, facilitating the flow of interdisciplinary concepts and methodologies. Notably, communication studies plays a pivotal role as a “knowledge converter,” establishing transmission pathways from electrical engineering through communication studies to interdisciplinary social sciences. As shown in

Table 6. The top 10 research fields in terms of citation frequency.

Citation Counts	Centrality	Research Fields
72	0.13	COMPUTER SCIENCE, THEORY and METHODS
56	0.09	COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS
52	0.12	COMPUTER SCIENCE, INFORMATION SYSTEMS
52	0.17	COMPUTER SCIENCE, CYBERNETICS
49	0.03	COMPUTER SCIENCE, SOFTWARE ENGINEERING
45	0.07	EDUCATION and EDUCATIONAL RESEARCH
42	0.13	COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE
42	0.19	ENGINEERING, ELECTRICAL and ELECTRONIC
27	0.01	TELECOMMUNICATIONS
23	0.09	ARCHITECTURE

, this highlights constrained knowledge exchange toward humanities-oriented disciplines and inefficient technology diffusion, with significant centrality disparities between artificial intelligence (0.13) and interdisciplinary social sciences (0.10)—a relative difference reaching 30%.

As shown in Figure 16, the total number of relevant studies in the field of art studies is 13, which fails to rank among the top in terms of proportion. This reflects a misalignment between its knowledge production model and the interdisciplinary paradigm, revealing its marginal structural position and developmental bottlenecks.

As shown in

Table 7. Top 9 research fields ranked by academic centrality.

Ranking	Subject Names	Centrality	Active Years
1	Electronic Engineering	0.19	2002
2	Cybernetics	0.17	1998
3	Communication Studies	0.16	2005
4	Urban Planning	0.15	2002
5	Computer Theory and Methods Artificial Intelligence	0.13	1998
6	Information System Management Science	0.13	1998
7	Environmental Studies	0.12	2000
8	Interdisciplinary Social Sciences	0.12	2006
9	Electronic Engineering	0.10	2014

, the centrality of the Electronic Engineering Center was 0.19 in 2002, and it dropped to 0.10 by 2014, showing a downward trend. The centrality of urban planning (0.15) and environmental studies (0.12) is noteworthy, reflecting a clear trend toward applied integration, though social science participation remains relatively limited. The data presented outlines an imperative multi-faceted strategy for the future of interactive space research.

4.3. Research Hotspots and Research Strategies

4.3.1. Keyword Co-Occurrence Network

Compared to co-citation analysis, keyword co-occurrence analysis can more intuitively display the hotspot content, thematic distribution, and disciplinary structure of interactive space to demonstrate the popular trends and knowledge structure effectively [119], revealing the main content, methods, and core ideas of the articles. In this study, the keyword co-occurrence map for the interactive space was generated. As shown in Figure 17, the number of nodes in the figure is 511, the number of links is 793, and the density is 0.0051. Node size represents the frequency of occurrence, and the link value represents the co-occurrence frequency of two keywords [120]. The thicker the line, the more frequent the occurrence.

As an indispensable visualization method within CiteSpace, the Keyword Mountain Graph provides a topographical representation of knowledge evolution. As generated and presented in Figure 18, this tool enables a structured parsing of keyword dynamics. It thereby reveals not only the enduring academic significance of core research concepts but also uncovers the deep, underlying patterns that have shaped the developmental trajectory of the interactive space field over time.

Currently, interactive space research exhibits a deep integration of technology-driven advancements and scenario-based applications: centered around “#0 interactive spaces,” the technological focus lies on “#2 augmented reality” and “#4 ubiquitous computing,” revealing a research mainstream centered on immersive and seamless access. Simultaneously, “#1 sustainability” and “#3 collaborative governance” reflect the field’s expansion into application scenarios such as smart cities, environmental sustainability, and social governance, emphasizing the social value and ethical dimensions of technology. Moreover, “#5 auditory feedback” and “#6 chopping” suggest natural interaction and multimodal feedback are becoming critical to user experience, while “#9 new media participation” and “#10 assessment methodologies” signal growing emphasis on public engagement and quantifiable outcomes.

4.3.2. Keyword Co-Occurrence Time Zone

As shown in Figure 19, the CiteSpace-based keyword co-occurrence timeline uses five-year time slices, with node size representing frequency within each period. Purple node borders indicate influence; thicker borders denote greater impact. Colours correspond to the top ten keyword clusters on the left. The horizontal axis spans the 1990–2025 time series, while the vertical axis corresponds to cluster labels. Each cluster’s duration and intensity indicate its rise, development, and decline trajectory.

As shown in Figure 20, From 2005 to 2015, there was an emergence of more targeted keywords such as “interactive spaces,” “augmented reality,” and “collaborative governance,” which signaled a shift in research focus towards humanities and social applications. Since 2015, the emergence of novel terms such as “sustainable development” and “evaluation methodology” has signaled a new phase of interdisciplinary integration and sustainable evolution.

4.3.3. Keyword Cluster Analysis

As shown in Figure 21, the keyword clustering network generated by the LLR algorithm based on CiteSpace automatically aggregates co-occurring keywords into thematic clusters, effectively identifying research structures and evaluating cluster quality. Covering 1990 to 2025, node size represents term frequency, with smaller cluster numbers containing more keywords. The modularity Q-value reached 0.8648, and the average silhouette value achieved 0.9129, both exceeding the 0.9 threshold. This indicates a robust clustering structure with strong internal consistency and reliable separability.

Current interaction space research remains centered on “#0 interaction space,” forming a triadic framework deeply integrating technology, society, and sustainability: with #4 ubiquitous computing and #2 augmented reality as technological foundations, interactions driven by #5 auditory feedback, #6 operational behavior segmentation, and #8 occupancy rate collectively point toward three core application directions: #3 collaborative governance, #9 new media participation, and #1 sustainability.

4.3.4. Research Trends

By analysing the high-frequency changes in each keyword within a certain period, further exploring the main content of the key literature in the research field becomes possible, thereby revealing the transformation of popular research topics and research patterns [121]. To explore changes in popular research topics during various periods from 1990 to 2025, this paper examines the strength of keyword occurrence in the top 25 subject terms. Burst terms represent shifts in research hotspots within a period. The high frequency of changes in subject terms across a large number of papers was evaluated. These bursts provide an intuitive view of when each research topic emerged and how its dynamics have evolved over time.

Figure 22 shows that research hotspots in different fields vary significantly across time periods. Before 2002, research was scarce, with no burst words. From 2003 to 2010, research on interactive spaces primarily aimed to develop a universal technological foundation, but it fell into the “efficiency-first” tool rationality trap, neglecting emotional and human needs. From 2011 to 2019, experience-based dimensions were introduced through design, active learning, and expectation confirmation models. However, the absence of social constraints resulted in inadequate technological democratization. Since 2020, research literature on interactive space has increased rapidly. The field has become more diverse and entered a stage of social deepening with emerging topics such as social media and politics. The study identified a 10-year generational time lag between technology clusters (e.g., ubiquitous computing) and humanistic clusters (e.g., emotion), which significantly quantifies the structural contradictions in interactive space research. Although social topics such as politics and feminism emerged after 2019, their burst strength is less than half that of ubiquitous computing.

5. Discussion

This study provides a systematic visualization analysis of interactive space research, revealing its developmental trajectory, hotspots, and trends. Based on these results, we discuss future directions, challenges, and paradigms to guide subsequent research. This study conducted a visual analysis of 450 documents in the interactive space field from the Web of Science Core Collection between 1990 and 2025 using CiteSpace. Nine types of knowledge maps were constructed, including author co-citation, institutional collaboration, and keyword co-occurrence. A three-dimensional knowledge structure comprising the technological basis, humanistic interface, and social application was identified.

5.1. Research Trends and Speculations on Future Development Directions

Current research on interactive spaces exhibits a pronounced technology-dominated character, with most literature focusing on engineering aspects such as algorithm optimization, sensing technologies, and system implementation. However, this technology-centric research paradigm is revealing its limitations [122,123,124,125], particularly when addressing complex social, emotional, and environmental demands [126]. Therefore, this study speculates that research and practice in interactive spaces will gradually shift from technology-driven approaches toward human-centric orientations, placing greater emphasis on affective computing, user experience design, and social inclusivity [127,128,129]. On the other hand, this study reveals a pronounced disciplinary imbalance in interactive space research. This fragmentation limits the comprehensive application effectiveness of interactive spaces in real-world scenarios. Therefore, future research urgently requires the construction of an integrated “technology–humanities–institutional” tripartite research framework to promote deep convergence across computer science, environmental behavioral science, public policy, design, and art. Only through interdisciplinary collaboration can the multidimensional challenges faced by interactive spaces within complex urban systems be addressed [130,131,132,133]. Furthermore, existing interactive space systems predominantly rely on predefined scenarios and rules, lacking dynamic responsiveness to unforeseen events such as public health crises or climate disasters. Future interactive spaces must develop enhanced adaptive and learning capabilities [134,135,136]. Through real-time data sensing, machine learning, and feedback mechanisms, they should dynamically adjust and optimize spatial functions, thereby strengthening their resilience against uncertainty.

5.2. Projections for Technology–Society Integration

Technologically, interactive spaces are evolving from “utilitarian tools” providing basic functions to “platform-based entities” supporting immersive experiences and social interactions [137,138]. This study projects that spatial computing, mixed reality (MR), and multimodal interaction (e.g., auditory and tactile feedback) [139,140,141,142] will form the core foundation of next-generation interactive spaces. These technologies not only enhance user immersion and engagement but also facilitate broader social interaction and cultural expression, positioning interactive spaces as vital bridges connecting the physical and virtual realms, as well as individuals and communities. Concurrently, as global sustainability agendas advance, interactive spaces will emerge as key vehicles for achieving carbon neutrality goals and enabling smart governance [143,144]. This study proposes introducing innovative assessment metrics like “Carbon Emotional Equivalents” (CEEs) to comprehensively evaluate the dual benefits of interactive spaces for environmental efficacy and human well-being [145,146,147]. Furthermore, integrating technologies such as blockchain and Decentralized Autonomous Organizations (DAOs) can drive innovation in citizen participation and collaborative governance models [148,149]. This facilitates a shift from “data providers” to “co-creators,” enhancing the democratic and inclusive nature of these spaces.

5.3. Methodology and Research Paradigm Speculation

Current interactive space research lacks a systematic, multidimensional evaluation system, making it difficult to comprehensively measure its integrated social, environmental, and technological benefits. Therefore, this study calls for the construction of a comprehensive evaluation framework integrating technical performance, user experience, social impact, and environmental sustainability to support more scientific and responsible innovation practices [150,151,152,153]. Particularly at the knowledge integration and institutional levels, this study further proposes leveraging conceptual technological approaches like the “spatiotemporal folding engine” to transform intangible resources—such as historical cultural heritage and local knowledge—into digital innovation assets [154,155,156,157]. This would enhance the cultural depth and spatiotemporal continuity of interactive spaces. Concurrently, strengthening relevant legislation and ethical guidelines is essential to ensure technological development aligns with societal values and ethical standards, thereby preventing misuse and social exclusion.

5.4. Limitations and Future Research Recommendations

This study primarily relies on English-language literature from the Web of Science Core Collection, potentially introducing language and database biases while overlooking valuable insights from non-English academic outputs and gray literature (e.g., policy reports, industry white papers). Future research should broaden data sources to incorporate multilingual and diverse document types, thereby achieving a more comprehensive research perspective. Additionally, some data used in this study extends to 2025, carrying a degree of foresight and speculative nature; its validity requires validation through subsequent empirical research. It is recommended that future research conduct more longitudinal studies, comparative case analyses, and field experiments—particularly strengthening investigations into local practices in regions like China, Southeast Asia, and Africa—to test the applicability and scalability of theoretical frameworks. This will advance interactive space research toward greater maturity and practical relevance.

This study uses visual analysis to track a paradigm shift in interactive space research from technology-centric to human-centric. The findings advocate for a tripartite technology–humanities–institutions framework to advance interdisciplinary collaboration and adaptive systems. Despite challenges like disciplinary imbalance, scarce evaluation systems, and limited data, future work must integrate multi-source data, pursue long-term empirical studies, and establish ethical guidelines to steer the field toward inclusive, sustainable, and intelligent spaces that enhance human well-being.

6. Conclusions

Employing CiteSpace analysis, this study reveals a paradigm shift in interactive space research from technology-driven to humanistic approaches, forming a technology–humanity–institution framework. The research has evolved through three phases (1990–2005, 2006–2013, 2014–2025) and comprises three core clusters: the technical infrastructure (with its emphasis on ubiquitous computing and evaluation), the interaction experience (covering augmented reality and multimodal interaction), and the value objectives (focusing on sustainability and collaborative governance). They have been designed to align with global goals and public engagement.

Future research requires interdisciplinary frameworks combining technology, humanities, and policy. Enhancing dynamic learning mechanisms will allow adaptive responses to environmental and emergency challenges. Green interaction technologies—such as low-power sensors and edge computing—should be developed alongside metrics like “Carbon Emotional Equivalent” (CEE) to evaluate sustainability. Finally, ethical governance must address data security, standardization, and inclusive co-governance for equitable digital–physical spaces.

Future research requires interdisciplinary frameworks integrating technology, behavior studies, and policy. Key directions include adaptive learning mechanisms for real-time response, green technologies such as low-power sensors and metrics like Carbon Emotional Equivalent (CEE), and ethical governance models for data security and inclusive co-governance. Through systematic bibliometric and visualization analysis, this study reveals a paradigm shift in interactive space research from technology-centric to human-centric approaches. As a novel human habitat deeply integrating physical and digital dimensions, interactive space aims not merely to demonstrate technological showmanship but to silently enhance human capabilities, nurture interpersonal relationships, and safeguard human dignity. It serves as a solid foundation for humanity’s figurative dwelling in the era of digital civilization.