Available via license: CC BY 4.0
Content may be subject to copyright.
Sensors 2025, 25, 906 https://doi.org/10.3390/s25030906
Article
Intersections of Big Data and IoT in Academic Publications:
A Topic Modeling Approach
Diana-Andreea Căuniac 1,2, Andreea-Alexandra Cîrnaru 1,2, Simona-Vasilica Oprea 1,* and Adela Bâra 1
1 Department of Economic Informatics and Cybernetics, Bucharest University of Economic Studies, no. 6
Piaţa Romană, 010374 Bucharest, Romania; diana.cauniac@csie.ase.ro (D.-A.C.);
alexandra.cirnaru@csie.ase.ro (A.-A.C.); bara.adela@ie.ase.ro (A.B.)
2 Doctoral School of Economic Informatics, Bucharest University of Economic Studies,
010374 Bucharest, Romania
* Correspondence: simona.oprea@csie.ase.ro
Abstract: As vast amounts of data are generated from various sources such as social me-
dia, sensors and online transactions, the analysis of Big Data oers organizations the abil-
ity to derive insights and make informed decisions. Simultaneously, IoT connects physical
devices, enabling real-time data collection and exchange that transforms interactions
within smart homes, cities and industries. The intersection of these elds is essential, lead-
ing to innovations such as predictive maintenance, real-time trac management and per-
sonalized solutions. Utilizing a dataset of 8159 publications sourced from the Web of Sci-
ence database, our research employs Natural Language Processing (NLP) techniques and
selective human validation to analyze abstracts, titles, keywords and other useful infor-
mation, uncovering key themes and trends in both Big Data and IoT research. Six topics
are extracted using Latent Dirichlet Allocation. In Topic 1, words like “system” and “en-
ergy” are among the most frequent, signaling that Topic 1 revolves around data systems
and IoT technologies, likely in the context of smart systems and energy-related applications.
Topic 2 focuses on the application of technologies, as indicated by terms such as “technolo-
gies”, “industry” and “research”. It deals with how IoT and related technologies are trans-
forming various industries. Topic 3 emphasizes terms like learning and research, indicat-
ing a focus on machine learning and IoT applications. It is oriented toward research involving
new methods and models in the IoT domain related to learning algorithms. Topic 4 high-
lights terms such as smart, suggesting a focus on smart technologies and systems. Topic 5
touches upon the role of digital chains and supply systems, suggesting an industrial focus
on digital transformation. Topic 6 focuses on technical aspects such as modeling, system per-
formance and prediction algorithms. It delves into the eciency of IoT networks with terms
like “accuracy”, “power” and “performance” standing out.
Keywords: IoT; big data; Latent Dirichlet Allocation (LDA); Natural Language Processing
(NLP); network; data
1. Introduction
The continuous ow of data from various sources, including social media interac-
tions, embedded sensors, and online transactions, represents what is now referred to as
Big Data. The scale of these data presents valuable opportunities for deeper insights, en-
abling businesses and organizations to make data-driven decisions. Meanwhile, the Inter-
net of Things (IoT) connects physical devices to the internet, allowing them to gather and
Academic Editor: Joaquin Ordieres
Received: 5 December 2024
Revised: 31 January 2025
Accepted: 31 January 2025
Published: 2 February 2025
Citation: Căuniac, D.-A.; Cîrnaru,
A.-A.; Oprea, S.-V.; Bâra, A.
Intersections of Big Data and IoT in
Academic Publications: A Topic
Modeling Approach. Sensors 2025,
25, 906. hps://doi.org/10.3390/
s25030906
Copyright: © 2025 by the authors.
Licensee MDPI, Basel, Swierland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Aribution (CC BY) license
(hps://creativecommons.org/license
s/by/4.0/).
Sensors 2025, 25, 906 2 of 38
transmit data in real time. It is reshaping interactions with technology, inuencing every-
thing from smart home automation and urban infrastructure to industrial processes and
technological solutions [1].
The convergence of Big Data and IoT is driving advancements in areas such as pre-
dictive maintenance in manufacturing, intelligent trac systems in urban environments
and personalized solutions (for instance, in healthcare). As IoT devices generate vast
quantities of data, Big Data analytics has an important role in processing and interpreting
this information, revealing paerns and insights that would otherwise remain undetected.
This integration is accelerating progress in elds such as autonomous transportation, en-
ergy optimization and industrial operations. As a result, research into Big Data and IoT
remains highly relevant for the next generation of intelligent systems and digital services
[2,3].
The interaction between Big Data and IoT is an essential driver of transformative in-
novations across smart homes, smart cities and industries (the blue nodes in Figure 1 con-
nected with IoT devices). By using IoT’s capability to collect and transmit data alongside
Big Data’s analytical power, this integration facilitates real-time decision making, person-
alized solutions and beer performance.
In smart homes, IoT-enabled devices such as thermostats, lighting systems and voice
assistants collect data on user behavior and preferences. Big Data analytics processes this
information to deliver personalized solutions, such as optimizing energy consumption
based on daily routines. Additionally, predictive maintenance is enhanced by IoT appli-
ances that transmit performance data to cloud-based platforms, enabling proactive
maintenance scheduling and minimizing downtime.
In smart cities, IoT and Big Data are integral to optimizing city infrastructure. Trac
management systems rely on IoT sensors and GPS-enabled devices to monitor congestion
and trac ow. Then, Big Data applications optimize trac signals, manage vehicle re-
routing and enhance mobility. Environmental monitoring systems also benet from this
synergy, using IoT to track air quality, noise pollution and water conditions, while Big
Data analytics identies pollution hotspots and informs mitigation strategies.
Industries are likewise undergoing signicant changes due to the combined impact
of Big Data and IoT. Predictive maintenance and asset management benet from IoT sen-
sors that track equipment performance with Big Data processing these insights to antici-
pate and prevent costly failures. Supply chain management is also enhanced, as the real-
time IoT tracking of shipments and inventory enables more accurate logistics planning
and demand forecasting. Moreover, industrial processes become more ecient with IoT
systems monitoring production ows and workforce productivity, while Big Data analyt-
ics identies ineciencies and improves operations.
The synergy between Big Data and IoT is graphically represented in Figure 1. The
blue nodes refer to the IoT devices at the smart cities, smart homes and industry level,
whereas the green nodes refer to the Big Data analytics for these entities: cities, homes,
industries.
Sensors 2025, 25, 906 3 of 38
Figure 1. Synergy between Big Data and IoT.
With thousands of publications released annually, understanding the core trends, in-
uential researchers and the direction of academic inquiry becomes increasingly challeng-
ing. Our analysis oers a structured approach to navigating this vast body of knowledge,
helping to identify key themes, trace the development of specic research areas and un-
cover the collaborative networks that drive innovation in the eld.
Traditional methods of analyzing research output, such as literature reviews, while
valuable, often fail to capture content-based connections between studies. This is where
the application of NLP provides advantages. By adopting NLP techniques, it becomes
possible to process and analyze larger text data: the abstracts, keywords and titles of aca-
demic papers, thus uncovering semantic paerns that are not immediately visible through
more conventional numerical metrics such as citation counts. NLP allows for the extrac-
tion of thematic clusters and the identication of evolving research trends [4,5].
Through NLP, our research aims to reveal the dominant topics within Big Data and
IoT research and also how these topics have developed over time. Moreover, it seeks to
highlight the researchers and institutions that are at the forefront of innovation in these
areas. By analyzing the language and content of the literature, this approach oers a nu-
anced perspective of the research landscape, identifying emerging trends and guiding fu-
ture scholarly eorts. Ultimately, this contributes to a clearer understanding of the current
state of research in Big Data and IoT and its potential future directions.
The objectives of our research focus on providing a clear understanding of the re-
search landscape in the elds of Big Data and IoT. Through a detailed analysis, our re-
search aims to accomplish several goals that contribute to understanding how research in
these areas has evolved and where it is heading.
• First, our research seeks to identify key trends within the literature by analyzing ab-
stracts, keywords, and titles from a large corpus of publications. This includes
Sensors 2025, 25, 906 4 of 38
detecting frequently explored topics and themes as well as tracking the distribution
of keywords over time.
• The research also focuses on six primary topics within the Big Data and IoT elds.
These topics have been analyzed to understand how they have evolved, showing
how some areas of research have gained importance while others have declined or
remained steady.
• Another goal is to examine collaboration paerns between authors, institutions, and
countries. By mapping co-authorship, the research identies contributors and insti-
tutions that play a central role in Big Data and IoT research, illustrating how
knowledge and innovation are shared across regions.
• Additionally, our research examines the co-occurrence of keywords in the literature
to identify clusters of related research areas and show how these areas are connected.
This helps to demonstrate the relationships between various topics and the interdis-
ciplinary nature of research in Big Data and IoT. In the end, this research provides an
overview of the academic landscape in these elds, helping to guide future research,
foster collaboration and identify emerging areas.
The motivation behind our research stems from the need to navigate and understand
the rapidly expanding body of knowledge in the elds of Big Data and IoT given their
transformative potential across industries and society. Exploring the interdisciplinary na-
ture of Big Data and IoT, the research emphasizes the interconnectedness of these elds
with other domains, such as energy management, industrial optimization and urban de-
velopment. This interdisciplinary perspective is essential for driving holistic solutions to
complex challenges. Overall, our motivation is to advance understanding, foster innova-
tion and provide actionable insights that can guide the future of research and application
in Big Data and IoT.
While prior reviews on Big Data and IoT exist, our research sets itself apart by lever-
aging Natural Language Processing (NLP) techniques and topic modeling, such as Latent
Dirichlet Allocation (LDA), to analyze a large dataset of 8159 publications. Our main mo-
tivation was to extract recent and relevant information from these publications on timely
concepts such as Big Data and IoT. This data-driven approach uncovers paerns and se-
mantic connections that traditional literature reviews often fail to identify. By extracting
and analyzing six distinct thematic topics, our research provides a structured understand-
ing of research trends, oering insights into these elds. Additionally, our research em-
phasizes the evolution of Big Data and IoT over time, showing how specic areas of study
have grown or shifted in emphasis. It highlights the interdisciplinary nature of the elds,
exploring connections between smart systems, energy applications, industrial transfor-
mations and technical aspects such as prediction algorithms and system performance. Our
research also maps collaboration paerns among researchers, institutions and countries,
oering valuable insights into the research ecosystem. Moreover, it identies emerging
technologies, such as edge computing, blockchain and AI, as key drivers of innovation in
these elds.
Our research makes several signicant contributions to the understanding of Big
Data and IoT:
• First, it identies six primary thematic topics using Latent Dirichlet Allocation (LDA)
modeling, revealing key areas such as data systems and IoT technologies in smart
systems and energy applications, IoT applications across industries, machine learn-
ing and IoT methods, smart technologies, industrial transformations and technical
aspects like system performance and prediction algorithms.
• It also tracks the evolution of research themes over time, highlighting how specic
areas, such as energy applications or industrial IoT, have grown in prominence or
shifted in focus. Furthermore, the research examines keyword co-occurrence to
Sensors 2025, 25, 906 5 of 38
uncover relationships between dierent research areas, showcasing the interdiscipli-
nary nature of Big Data and IoT. This analysis identies clusters of related topics,
providing insights into the connections between diverse areas of study.
• A notable contribution is the identication of emerging technologies, including edge
computing, blockchain, and AI, which are driving innovation in Big Data and IoT.
The study sheds light on how these technologies contribute to smart systems, predic-
tive maintenance, industrial optimization, and other transformative applications.
• By uncovering semantic paerns and connections that are often missed in traditional
literature reviews, it oers a data-driven approach to understanding research trends.
This approach enhances the ability to navigate and interpret the rapidly expanding
body of knowledge in these elds. It emphasizes the interdisciplinary connections
between Big Data and IoT, particularly in areas such as smart systems, energy man-
agement, industrial operations, and urban development.
The organization of this paper follows a systematic structure. It begins with an intro-
duction (Section 1) that establishes the synergy between Big Data and IoT, emphasizing
their transformative potential across smart homes, cities and industries. This section high-
lights the research gap, detailing the absence of integrated bibliometric and topic model-
ing analyses in existing literature. It further outlines the objectives and motivation of the
study, seing the stage for the subsequent sections. Section 2 delves into the literature
review, which provides an extensive examination of prior research in the elds of Big Data
and IoT. This includes bibliometric studies focused on their applications in healthcare,
agriculture and industrial IoT. The methodology (Section 3) describes the research process
in detail, beginning with data collection from the Web of Science database, which resulted
in a dataset of 8159 publications. It outlines the data preprocessing steps, including clean-
ing, tokenization and lemmatization, which is followed by sentiment analysis techniques.
The use of Latent Dirichlet Allocation (LDA) for topic modeling is explained, along with
visualization tools like pyLDAvis, which are employed to analyze the relationships be-
tween topics and uncover hidden thematic structures.
In the results (Section 4), the paper presents a detailed analysis of publication trends,
such as the dominance of original research articles and the leading journals in the eld.
The outcomes of topic modeling are presented, revealing six key themes, including smart
systems, IoT applications and advancements in Big Data technologies. The conclusions
(Section 5) synthesize the key ndings. Additionally, the potential for exploring emerging
technologies like blockchain and edge computing is discussed.
2. Literature Review
2.1. Big Data and IoT
The synergy between IoT and Big Data is foundational to the advancement of smart
systems across various domains. IoT generates vast amounts of real-time data through its
interconnected devices—sensors, wearables, smart meters, and other IoT technologies—
across industries such as healthcare, transportation, smart cities, and manufacturing. In
essence, IoT serves as the data source, continuously producing information that needs to
be stored, processed, and analyzed eciently. Big Data technologies, such as Hadoop,
Spark, and cloud computing, provide the infrastructure and analytical models to handle
this volume, velocity, and variety of data. This synergy allows organizations to achieve
smarter, more ecient operations, improve customer experiences, and innovate with ad-
vanced features like real-time analytics, predictive maintenance, and automation.
The article “Latest advancements and prospects in the next-generation of Internet of
Things technologies” contributes to this understanding by emphasizing the critical role of
Big Data analytics in IoT ecosystems [6]. The authors highlight how Big Data can be
Sensors 2025, 25, 906 6 of 38
leveraged to extract valuable insights from the extensive data generated by IoT devices.
By analyzing paerns, trends, and correlations in the IoT-generated data, Big Data tech-
nologies enhance decision-making processes, support predictive models, and enable the
creation of personalized services tailored to individual or organizational needs. The paper
underlines that this combination not only optimizes IoT system performance but also fos-
ters the development of more intelligent and autonomous systems.
Moreover, the article delves into the challenges faced by IoT systems, particularly
concerning data management, security, and privacy. Big Data tools oer solutions to these
challenges by providing scalable data storage, real-time processing capabilities, and ad-
vanced analytics that improve the eciency of IoT systems. The paper underscores that
the next generation of IoT technologies will increasingly rely on the synergy between IoT
and Big Data to overcome existing limitations and unlock new potential in smart applica-
tions. Through this synergistic relationship, the capabilities of IoT can be maximized, en-
suring that the massive ow of data is harnessed eectively to drive intelligent outcomes.
The rapid advancement of Big Data and the Internet of Things (IoT) has generated
extensive academic interest, which is reected in a signicant number of publications. Re-
cent bibliometric analyses oer a systematic view of the research landscape in these elds,
identifying key themes, methodologies, and research gaps. For instance, analyses con-
ducted by MDPI and IEEE Xplore have highlighted several recurring themes, including
real-time data processing, smart city applications and healthcare IoT implementations [7].
Bibliometric studies reveal that collaboration in Big Data and IoT research has ex-
panded signicantly over recent years, with co-authorship networks centered around
countries with substantial technological infrastructures, such as the United States and
China. Research clusters often focus on applications in smart cities, healthcare and indus-
trial IoT, where real-time analytics and data integration are paramount challenges.
Tools like VOSviewer and SciMAT are frequently used to map author collaboration,
topic co-occurrences, and research productivity, helping to visualize the eld’s evolution
and highlight prolic authors and institutions. For example, Özköse, in his study “Biblio-
metric Analysis and Scientic Mapping of IoT”, used SciMAT as a tool for understanding
the existing landscape and potential growth in IoT research [8].
The article “Exploring Machine Learning: A Scientometrics Approach Using Biblio-
metrix and VOSviewer” analyzes the machine learning research landscape through bibli-
ometric methods. The authors used VOSviewer and Bibliometrix to visualize trends in
publications, identify inuential authors, and uncover key research themes. They present
some signicant contributions in the eld and illustrate collaboration networks among
researchers, as well as the increasing relevance of machine learning across various disci-
plines, suggesting future research directions [9].
Goranin et al. in their article “A Bibliometric Review of Intrusion Detection Research
in IoT” oer a comprehensive look at the evolution of intrusion detection systems (IDSs)
in IoT networks [10]. The methodology in the article involved a bibliometric analysis to
examine intrusion detection research trends in IoT. The authors collected data from the
Web of Science database, focusing on research published from 2017 to 2023. They used
statistical and network analysis tools to assess publication volume, keyword co-occur-
rences, citation paerns, and collaboration networks among researchers and institutions.
Visual mapping software, such as VOSviewer, helped visualize trends and emerging ar-
eas. By leveraging tools like VOSviewer, the authors identify high-impact research areas
such as machine learning applications in IDSs, challenges around data security and pri-
vacy, and the growing importance of real-time detection systems in IoT. The review sug-
gests research directions, including enhancing algorithmic eciency and increasing inter-
national collaboration to address IoT security challenges. This approach provided insight
Sensors 2025, 25, 906 7 of 38
into high-impact topics, key contributors, and collaborative eorts in IoT intrusion detec-
tion research [10].
Parlina et al. examined trends in Big Data research from 2009 to 2018, focusing on
topics, collaboration networks, and keyword analysis using data from Scopus [11]. Re-
searchers used co-word analysis and VOSviewer to map topics, identifying signicant
themes like machine learning, data analytics, and management. They found Big Data re-
search increasingly intersects with elds like healthcare, nance, and social sciences, con-
cluding that its interdisciplinarity reects growing practical and theoretical complexity.
The authors Abdullahi et al. conducted a bibliometric analysis to assess the evolution
of Internet of Things (IoT) applications in smart agriculture [12]. They utilized data from
the Scopus database covering publications from 2012 to 2022. They employed co-citation
analysis and keyword co-occurrence methods, leveraging VOSviewer software to visual-
ize trends, collaborations, and thematic clusters in the research landscape. The analysis
revealed a signicant increase in publications related to IoT applications in agriculture,
particularly from 2017 onwards. Key themes identied include precision agriculture, re-
mote sensing, and smart farming technologies. The study highlights strong collaborations
among institutions with notable contributions from regions such as Europe and North
America. The authors suggest that future research should focus on developing more inte-
grated IoT systems to address challenges in smart agriculture, emphasizing sustainability
and resource eciency.
Emerging trends in the literature focus on specic technological and ethical chal-
lenges. For instance, while advancements in IoT hardware and real-time processing capa-
bilities are regularly highlighted, studies also emphasize the ethical implications sur-
rounding data privacy and security. Notably, the integration of Big Data analytics with
IoT technologies introduces new layers of complexity in managing vast data volumes,
particularly when sensitive information is involved in domains like healthcare and urban
planning [13,14].
Despite signicant advances, bibliometric analyses identify critical gaps that future
research could address. Areas requiring deeper exploration include enhanced frame-
works for data interoperability across IoT platforms and advanced machine learning al-
gorithms to improve predictive capabilities within Big Data environments [15].
Additionally, there is a noted need for studies focusing on scalable architectures that
support the growing data inux in IoT systems. As the eld matures, further bibliometric
studies may continue to map the evolving intersections of Big Data and IoT, providing
actionable insights for both academia and industry.
Another prominent eld is smart city development, where IoT and Big Data are es-
sential for managing urban resources and infrastructure. In their bibliometric analysis,
Alaeddini et al. identied primary themes around energy management, trac optimiza-
tion, and environmental monitoring. The authors used cluster analysis to show how Big
Data technologies enable smart city frameworks to make data-driven decisions. They sug-
gest that future studies focus on integrating machine learning with IoT in real time to
address the complex demands of urban populations. These technologies not only help in
monitoring but also provide predictive insights that can aid in urban planning [16].
The article “An Edge-Fog-Cloud computing architecture for IoT and smart metering
data” introduces a comprehensive computing architecture that integrates Edge, Fog, and
Cloud computing to process data from IoT devices and smart metering systems [17]. The
authors propose a framework that distributes data processing tasks across these three lay-
ers to enhance eciency and reduce latency. By allocating specic tasks to each layer—
Edge, Fog, and Cloud—the architecture aims to optimize resource utilization and improve
the overall performance of IoT applications.
Sensors 2025, 25, 906 8 of 38
The paper emphasizes the importance of this layered approach in managing the di-
verse and voluminous data generated by IoT devices and smart meters. The Edge layer
handles immediate data processing needs, the Fog layer manages intermediate processing
and storage, and the Cloud layer provides extensive computational resources for complex
analytics and long-term data storage. This hierarchical structure ensures that data are pro-
cessed at the most appropriate level, balancing the trade-os between speed, resource
consumption, and computational power.
In “A bibliometric analysis of IoT applications in logistics and supply chain manage-
ment”, the authors delve into the integration of IoT within the logistics and supply chain
management sector [18]. This study provides a comprehensive bibliometric overview, ex-
ploring how IoT technologies optimize processes in logistics and supply chains, particu-
larly focusing on Big Data analytics and IoT-driven innovation. The paper emphasizes the
potential of IoT to enhance the food supply chain, highlighting how data from IoT devices
can be used for beer resource management and more ecient operations. Through cita-
tion analysis and trends, the study captures the rapid evolution of these technologies, of-
fering insights into the growing importance of IoT in transforming supply chain dynam-
ics.
The article “Education big data and learning analytics: a bibliometric analysis” ana-
lyzes trends in the intersection of Big Data and education, specically focusing on learning
analytics [19]. It presents a thorough examination of how Big Data are being utilized in
educational seings to track learning paerns and improve student outcomes. The paper
not only reviews the existing literature but also provides a forward-looking perspective,
oering recommendations for the future of learning analytics. The authors point out the
potential for Big Data to revolutionize personalized learning and enhance pedagogical
methods, encouraging further research to rene educational practices and technologies.
Lars Lundberg, in his study, oers an in-depth exploration of the ever-evolving Big
Data eld, aiming at identifying key research trends and future directions in a rapidly
growing domain [20]. He used bibliometric methods to analyze the vast body of literature
related to Big Data, uncovering critical themes such as data processing, storage, and pri-
vacy concerns. The author also touched on emerging areas like articial intelligence and
machine learning, emphasizing the interdisciplinary nature of Big Data research. This pa-
per is pivotal for researchers seeking to understand the broad scope of Big Data and its
implications for various sectors.
The paper “A Bibliometric Analysis of Research on Big Data and Its Potential to Value
Creation and Capture” focuses on the scientic landscape of Big Data specically in terms
of its potential for value creation and capture [21]. The study applies bibliometric tech-
niques to assess how Big Data can drive economic growth and business value. By review-
ing academic literature on the subject, the authors outline the evolving role of Big Data in
shaping business strategies, operational eciencies, and innovations across industries.
This analysis is particularly insightful for understanding the relationship between Big
Data and organizational value creation, oering valuable insights for both scholars and
industry professionals.
In “Visualizing the Landscape of Home IoT Research: A Bibliometric Analysis”, the
authors provide a detailed overview of the research landscape surrounding home IoT.
Bibliometric methods were used to map the key trends, challenges, and future directions
in the domain of smart home technologies. They focus on the various IoT devices and
systems used in homes, from security solutions to energy management systems, and ex-
plore how these technologies are shaping modern living environments. The study identi-
es gaps in research, particularly in areas such as user privacy, device interoperability,
and the integration of AI with IoT in the home context [22].
Sensors 2025, 25, 906 9 of 38
Wang et al., in the reference [23], examine the role of edge computing in the IoT eco-
system, focusing on how it facilitates the processing of Big Data generated by IoT devices.
This paper provides valuable insights into the convergence of edge computing and IoT,
detailing how edge computing can enhance data processing speed, reduce latency, and
improve the eciency of IoT networks. Through a bibliometric analysis, the study high-
lights the rapid advancements in edge computing technologies and the growing interest
in this eld within IoT research.
Each of these studies uses bibliometric methods to analyze vast domains of research,
oering a critical examination of trends, emerging topics, and gaps in their respective
elds. From logistics and education to IoT and Big Data, these articles highlight the inter-
connectedness of technological advances and their potential to shape industries in pro-
found ways.
Collectively, these studies point to several challenges in Big Data and IoT integration,
notably in data security, interoperability, and the need for scalable infrastructure. As IoT
ecosystems grow, managing the sheer volume and complexity of data requires advance-
ments in real-time analytics and machine learning. Ge et al. present in their article titled
“Big Data in Internet of Things: A Survey” a comprehensive overview of how Big Data
are utilized within IoT ecosystems. It outlines the unique challenges presented by the mas-
sive volumes of data generated by IoT devices, including data storage, processing and
analysis [24]. The authors explore various architectures and frameworks designed to fa-
cilitate eective Big Data management in IoT contexts. Additionally, the survey highlights
potential applications, future trends and the need for advanced analytical techniques and
security measures.
Each domain presents unique requirements, whether in patient privacy in healthcare
or real-time reliability in industrial seings [25]. Future research may thus benet from
cross-disciplinary approaches that address these foundational challenges while continu-
ing to advance the specic needs of each domain.
In an extensive bibliometric analysis on Big Data and IoT in smart infrastructure, Guo
et al. used data from Web of Science to map publication trends, highlighting rapid growth
in areas like real-time monitoring, data visualization, and smart grids [26]. Their study
identied clusters around infrastructure resilience, energy eciency, and smart transpor-
tation, underscoring the role of Big Data in enhancing IoT-driven urban management. The
authors also found that collaboration among industry and academia has increased, par-
ticularly in regions investing in smart city initiatives.
The adoption of IoT and Big Data in healthcare has been examined by Šajnović and
her co-authors. This analysis focused on wearable technologies and predictive analytics
for patient care, showing a surge in publications related to chronic disease management
and real-time health monitoring. According to them, Big Data enable IoT devices to gen-
erate insights for personalized healthcare, though challenges such as data privacy and
interoperability remain prominent. Their analysis reveals the importance of data govern-
ance in healthcare IoT, advocating for secure, interoperable systems to manage sensitive
health data eectively [27].
IoT and Big Data have also seen signicant application in agriculture, as evidenced
by a bibliometric study by Misra et al. This research highlights IoT’s role in precision ag-
riculture, where real-time data on soil conditions, crop health, and weather forecasts ena-
ble farmers to optimize resources. The authors note a growth in publications on environ-
mental sustainability and resource conservation with keywords such as “climate adapta-
tion” frequently appearing. They call for future research on robust analytics frameworks
that can process large, diverse agricultural data to address global food security concerns
[28].
Sensors 2025, 25, 906 10 of 38
In the industrial sector, IoT applications for predictive maintenance and asset man-
agement have been a focal point, as explored in a bibliometric study by Keleko et al. Their
analysis reveals a rising interest in IoT applications for supply chain optimization and
automated quality control. Big Data analytics allows organizations to anticipate equip-
ment failures and optimize production through predictive maintenance, underscoring the
need for robust data management strategies and the integration of advanced analytics.
The authors advocate for future research to address challenges related to implementation
and scalability [29].
IoT plays an important role in the educational sector as well. The article “A Compar-
ative Study of Chinese and Foreign Research on the Internet of Things in Education: Bib-
liometric Analysis and Visualization” analyzes publications related to IoT in education,
comparing trends between Chinese and international research. The study reveals distinct
research focuses with China emphasizing smart campus implementations while foreign
research tends to explore broader applications. It concludes by suggesting areas for future
research and collaboration to enhance the integration of IoT in educational contexts [30].
2.2. Smart Industries
The intersection of Big Data and IoT technologies continues to aract signicant
scholarly interest across diverse domains, including agriculture, industrial management,
and business. A review of recent bibliometric analyses highlights evolving research trends
and emerging thematic clusters within these elds [31].
For instance, IoT’s role in smart agriculture has been extensively analyzed by Liang
and Shah, who mapped out core themes such as precision agriculture, climate change ad-
aptation, and smart irrigation. The study emphasizes that IoT adoption in agriculture of-
ten leverages wireless sensor networks and machine learning to optimize water usage and
soil management, ultimately addressing food security challenges. The authors identied
a dramatic increase in IoT-related publications in agriculture, with topics like machine
learning and smart irrigation showing high growth rates, indicating both scholarly and
practical interest in enhancing agricultural productivity [32].
Research shows the growing role of IoT in agriculture, emphasizing its potential to
address food security and environmental challenges by integrating advanced data-driven
technologies such as blockchain, UAVs, and IoT networks to monitor soil, crops, and re-
source use eciently. Studies on Industry 4.0 technologies in agriculture reveal that IoT
and Big Data analytics are crucial for managing vast agricultural data volumes, helping
to optimize crop yields and reduce environmental impact [33].
Similarly, a bibliometric study by researchers in the eld of industrial IoT (IIoT) ex-
plores IoT applications in business management and operational eciency. According to
recent analyses, IIoT’s adoption within manufacturing has been pivotal with applications
in automated production, industrial safety, and smart maintenance. Such research has fre-
quently targeted the enhancement of industrial processes through IoT-driven automation
and asset management, illustrating the operational impact of IoT beyond traditional IT
applications. Key ndings suggest a need for further research on interoperability and cy-
bersecurity within industrial IoT systems [34,35].
On the business side, IoT has been instrumental in creating new business models and
customer engagement strategies. Bibliometric analyses show that scholars are increas-
ingly interested in IoT as a driver of innovation from smart consumer products to service
personalization. Researchers have mapped trends in IoT’s role in customer proling, se-
cure transaction processing, and digital wallets, emphasizing IoT’s transformative poten-
tial in areas such as nance and consumer electronics [36].
A comprehensive bibliometric analysis focusing on IoT in environmental sustaina-
bility reveals a strong emphasis on IoT applications in resource conservation and climate
Sensors 2025, 25, 906 11 of 38
monitoring. IoT technologies are seen as vital for enhancing sustainability practices
through energy management, water conservation, and real-time environmental monitor-
ing. This area remains a dynamic research eld with publications rapidly increasing as
governments and industries invest in IoT for environmental resilience [37]. These studies
collectively underscore IoT’s versatility across various domains. They also highlight chal-
lenges related to data integration, privacy, and system interoperability that continue to
shape research directions in Big Data and IoT technologies.
The article “Anomaly Detection with Machine Learning Algorithms and Big Data in
Electricity Consumption” explores the application of machine learning (ML) techniques
to identify anomalies in electricity usage data [38]. In this study, the authors analyze large
datasets of readings provided by smart meters installed in a trial study in Ireland. They
apply a hybrid approach that combines various machine learning algorithms to detect
anomalies in electricity consumption paerns. The research aims to enhance the accuracy
and eciency of anomaly detection systems, which is crucial for identifying fraudulent
activities, equipment malfunctions, or other irregularities in electricity usage.
By leveraging Big Data analytics, the study addresses the challenges associated with
processing and analyzing large volumes of electricity consumption data. The integration
of machine learning algorithms with Big Data enables the development of more robust
and scalable anomaly detection systems, leading to improved energy management and
reduced operational costs.
In a bibliometric study focused on IoT and Big Data in healthcare, Dian, Vahidnia
and Rahmati identied increasing interest in wearable devices, predictive analytics, and
patient monitoring. Their analysis revealed clusters of research around chronic disease
management and real-time data monitoring systems. This growth is driven by the need
for continuous patient data to support preventative healthcare and timely intervention,
where Big Data analytics help process the extensive data streams generated by IoT de-
vices. This study highlights key challenges, including data security and interoperability
among healthcare IoT devices, as pressing issues for future research [39].
IoT and Big Data applications in agriculture and environmental monitoring have
gained aention, as shown in a 2023 bibliometric analysis by Pachouri et al. This study
identies IoT’s use in precision agriculture for resource conservation, especially in water
management and soil monitoring. Big Data analytics are essential for making data-driven
decisions, where IoT sensors track crop health and environmental conditions in real time.
The authors found that as climate change concerns grow, so does research in this area
with an emphasis on using IoT for sustainable agriculture practices [40,41].
2.3. Smart Homes
The convergence of IoT, Big Data, AI and sustainable technologies has signicantly
impacted the development of smart homes and cities. Smart cities aim to enhance urban
eciency and sustainability, employing IoT and wireless sensor networks for resource
optimization and infrastructure management [42]. In particular, urban water manage-
ment strategies, leveraging smart technologies, address critical issues like water conser-
vation and distribution [43]. These advancements contribute to meeting global sustaina-
bility goals and improving city resilience.
Smart homes are transforming residential spaces by integrating intelligent systems
tailored for specic user needs, particularly older adults. Bibliometric analyses reveal a
focus on assistive technologies for health monitoring, fall detection and personalized care
[44,45]. In reference [46], the authors map a decade of research on smart homes for the
elderly with a specic focus on studies indexed in Web of Science. They use CiteSpace for
scientometric analysis, emphasizing co-citation paerns and keyword clustering. This re-
search takes a longitudinal approach, analyzing how the research on smart homes for
Sensors 2025, 25, 906 12 of 38
elderly individuals has evolved over the past decade. The authors identify shifts in re-
search focus, from basic technology implementation to the integration of health monitor-
ing, safety and comfort features. The use of CiteSpace is integral in visualizing these shifts
and providing a comprehensive overview of the trends in smart homes for the elderly.
Innovative designs in smart buildings emphasize safety and comfort, addressing
risks for vulnerable populations, as depicted in [47]. This research used a systematic and
bibliometric analysis to explore the development of smart buildings that reduce indoor
risks for safety and health, particularly for the elderly. The authors combine citation anal-
ysis with thematic clustering, using tools such as VOSviewer and CiteSpace to map out
the evolution of smart building technologies. They analyze the intersection of smart tech-
nologies, building safety and elderly health. This combination of bibliometrics and sys-
tematic review helps the authors highlight emerging trends in smart building design and
identify gaps in research, especially related to safety risks and environmental factors af-
fecting elderly residents. Furthermore, IoT-driven solutions are revolutionizing indoor air
quality monitoring, showcasing the potential of real-time data collection for environmen-
tal health [46].
Security and privacy remain prominent concerns, as highlighted in research explor-
ing IoT frameworks for smart homes. Eective strategies to mitigate risks are necessary to
foster user trust and system adoption [48]. In this research, the authors discuss various
security protocols and solutions that have been proposed in the literature to safeguard
smart home systems. Their ideas contribute to the growing body of knowledge on cyber-
security in IoT-enabled environments, emphasizing the importance of securing personal
data and preventing unauthorized access. Moreover, systematic reviews of trends empha-
size the importance of interdisciplinary research, suggesting that future advancements
will hinge on collaborative eorts between technologists and urban planners [49,50].
2.4. Smart Cities
Smart cities are rapidly evolving into key areas of interdisciplinary research that span
across domains like sustainability, technology, governance, urban planning and educa-
tion. Bibliometric analyses have become an essential tool to systematically map the grow-
ing body of knowledge in smart city research. In the past few years, a number of studies
have employed various bibliometric methods to provide insights into trends, collabora-
tions and emerging areas of study. For instance, the research by Guo et al. provides a
broad overview of smart city research through an extensive bibliometric analysis using
VOSviewer and CiteSpace . This research analyzes over 4000 papers to map the co-author-
ship networks, identify key research clusters and examine the evolution of smart city top-
ics. The authors focus on the identication of emerging trends over a span of two decades,
providing a longitudinal perspective on the growth of the eld. Their method primarily
utilizes co-authorship and keyword co-occurrence to identify and visualize the intellec-
tual landscape of smart cities. By mapping collaborative relationships and the frequency
of specic terms, they oer a comprehensive view of how interdisciplinary collaborations
have shaped smart city research. It serves as a foundational work for understanding the
broader scope of smart city research and its global development.
In contrast, Scala et al. take a more niche approach by exploring the conceptualization
of smart cities specically in the context of education [51]. Their bibliometric study em-
ploys co-word analysis and thematic clustering techniques to map how the concept of
education has been integrated into the broader smart city paradigm. The study focuses on
publications from Scopus and Web of Science, employing keyword co-occurrence analysis
to identify the dominant themes in smart city education. This approach allows them to
visualize the evolution of educational concepts within smart cities and how they intersect
with urban technologies. By employing thematic mapping, Scala et al. oer a fresh
Sensors 2025, 25, 906 13 of 38
perspective, focusing on the role of education in urban environments and how it contrib-
utes to the development of smart cities. This specialized focus on education sets their
study apart from the broader, more general studies in the eld.
Rejeb et al. present an innovative combination of bibliometric analysis and main path
analysis to trace the intellectual development of smart city research [52]. While biblio-
metric methods such as citation and co-citation analysis are used to map out key articles
and authors, the main path analysis adds an additional layer of sophistication by identi-
fying critical research trajectories that have shaped the evolution of the eld. The main
path method focuses on citation chains to reveal how foundational studies have inu-
enced the direction of subsequent research. This dual approach of bibliometric mapping
and main path analysis is particularly valuable for understanding the underlying path-
ways of knowledge ow in smart city research.
In a more focused area of smart city technology, Gupta et al. examine the intersection
of AI and smart cities [53]. Their bibliometric analysis leverages citation analysis and key-
word co-occurrence to uncover trends related to the use of AI in urban management. By
employing bibliographic coupling, the study identies the relationship between seminal
papers and newer publications, providing insights into how AI is becoming integrated
into smart city frameworks. Although the scope of the study is narrower in comparison
to Guo et al.’s, it provides valuable insight into the increasing role of AI technologies in
enhancing urban eciency, sustainability, and overall city management [26,53]. The
study’s specic focus on AI as a transformative tool for smart cities oers a perspective in
the rapidly evolving eld of AI applications in urban environments.
The primary goal of Hajoary et al.’s study is to map the research landscape, high-
lighting the most prolic researchers, institutions, and countries contributing to smart city
scholarship. Additionally, it examines the thematic evolution of the eld, identies col-
laboration networks within the research community, and pinpoints emerging areas of fo-
cus [54]. While the study’s strength lies in its comprehensive and methodologically rigor-
ous approach, it does have some limitations. The exclusive focus on English-language
publications may exclude signicant contributions in other languages, and its academic
orientation does not encompass practical insights available from industry reports or pa-
tents.
Puliga et al. introduce an innovative methodological approach by combining a bibli-
ometric analysis of scientic publications with patent analysis [55]. While many biblio-
metric studies focus solely on academic papers, like the article from reference [54], these
authors expand the scope to include patents, oering a more holistic view of smart city
innovations. By utilizing tools such as VantagePoint, they analyze both scientic publica-
tions and patent trends to uncover research and technological developments in the smart
city sector. This approach is particularly valuable as it captures the practical, industry-
driven aspects of smart cities alongside theoretical research. Their study emphasizes the
importance of standardization and interoperability challenges in the smart city sector,
shedding light on the industrial applications that stem from academic research.
A brief comparison of previous studies is provided in Table 1.
Table 1. Comparison of previous studies.
Ref
Objectives
Methods
Keywords Focus
Main Findings
[6]
Overview of advancements in IoT, includ-
ing complex networks, SIoT, and Big Data
analytics.
Bibliometric analysis focus-
ing on new IoT technolo-
gies and integration of SIoT
and Big Data.
IoT, Big Data, social
IoT, complex net-
works, research
trends.
Highlights emerging technologies like SIoT
and Big Data analytics, and the future of IoT in
optimizing data processing and enhancing au-
tomation.
[7]
Highlight recurring themes in IoT and Big
Data research.
Literature review of MDPI
and IEEE Xplore analyses.
IoT, Big Data, recur-
ring themes.
Recurring themes include real-time data pro-
cessing and smart applications.
Sensors 2025, 25, 906 14 of 38
[8]
Scientic mapping of IoT research land-
scape.
Bibliometric analysis using
SciMAT.
IoT, research land-
scape, SciMAT.
Identied growth areas and potential in IoT re-
search.
[9]
Analyze machine learning research trends
using bibliometrics.
Bibliometric methods using
VOSviewer and Biblio-
metrix.
Machine learning,
trends, collaboration.
Visualized trends and inuential authors in
machine learning.
[10]
Review trends in intrusion detection sys-
tems for IoT.
Bibliometric analysis and
network analysis.
Intrusion detection,
IoT, network analy-
sis.
Highlighted trends in IoT security and ma-
chine learning for IDS.
[11]
Identify themes, collaborations in Big Data
research.
Co-word analysis and
VOSviewer mapping.
Big Data, themes, col-
laborations.
Big Data intersect with multiple domains like
healthcare and nance.
[12]
Assess IoT applications in smart agricul-
ture.
Co-citation and keyword
co-occurrence analysis.
Smart agriculture,
IoT, sustainability.
IoT applications in agriculture have grown sig-
nicantly since 2017.
[13],
[14]
Discuss ethical implications of IoT and Big
Data.
Ethical discussion and the-
matic review.
Ethical implications,
privacy, security.
IoT integration raises ethical and data security
concerns.
[15]
Identify research gaps in IoT interoperabil-
ity and ML.
Bibliometric analysis of re-
search gaps.
Interoperability, ma-
chine learning.
Research gaps in interoperability and ad-
vanced ML frameworks.
[16]
Study IoT and Big Data in smart city devel-
opment.
Cluster analysis of publica-
tion trends.
Smart cities, energy,
trac optimization.
Big Data and IoT improve urban planning and
resource allocation.
[17]
Proposed an Edge–Fog–Cloud architecture
to process IoT and smart metering data.
Hybrid architecture com-
bining edge, fog, and cloud
computing layers.
IoT, Edge computing,
Fog computing,
Cloud computing,
smart metering.
Enhances system eciency and reduces la-
tency by distributing processing tasks across
Edge, Fog, and Cloud layers.
[18]
Examines the integration of IoT in logistics
and supply chain management, focusing on
innovation and optimization.
Bibliometric analysis of
published IoT research in
logistics.
IoT, logistics, supply
chain management,
Big Data optimiza-
tion.
Highlights key areas of innovation in IoT for
logistics, particularly big data optimization and
food supply chain applications.
[19]
Investigates trends in education Big Data
and learning analytics.
Bibliometric analysis of ed-
ucation-related Big Data
and analytics.
Education, Big Data,
learning analytics.
Provides trends and recommendations for im-
proving educational outcomes using Big Data
and analytics.
[20]
Identies research directions and trends in
the fast-evolving Big Data eld.
Bibliometric mining of Big
Data research.
Big Data, research
trends, bibliometric
analysis.
Describes emerging research trends and oppor-
tunities within the Big Data eld.
[21]
Analyzes Big Data research related to value
creation and capture.
Bibliometric analysis focus-
ing on Big Data’s potential
for business value.
Big Data, value crea-
tion, value capture.
Assesses the impact of Big Data on value crea-
tion and business strategies.
[22]
Visualizes research trends in the eld of
home IoT.
Bibliometric analysis using
VOSviewer for trend visu-
alization.
Home IoT, smart
homes, IoT research
trends.
Provides insights into key trends and future di-
rections in home IoT research.
[23]
Explores research on edge computing appli-
cations in IoT.
Bibliometric analysis of
edge computing research.
Edge computing, IoT,
security, privacy.
Highlights the critical role of edge computing
in IoT systems, focusing on security, trust, and
privacy.
[24]
Map IoT and Big Data in smart infrastruc-
ture.
Mapping publication
trends using Web of Sci-
ence.
Smart infrastructure,
IoT, grids.
IoT enhances smart grids and urban resilience.
[25]
Survey Big Data challenges in IoT ecosys-
tems.
Survey of Big Data manage-
ment frameworks.
Big Data, IoT, archi-
tectures.
Big Data frameworks address IoT scalability
and security.
[26]
Highlight cross-disciplinary challenges in
IoT and Big Data.
Thematic analysis of inter-
disciplinary challenges.
Cross-disciplinary
IoT challenges.
IoT research requires cross-disciplinary solu-
tions.
[27]
Study IoT in healthcare, wearable technol-
ogy focus.
Bibliometric review of
wearable IoT.
Healthcare IoT, wear-
able technology.
Wearable IoT devices support chronic disease
management.
[28]
Analyze IoT in precision agriculture for
sustainability.
Keyword analysis and bib-
liometric methods.
Precision agriculture,
climate, IoT.
IoT optimizes agricultural practices and cli-
mate adaptation.
[29]
Explore predictive maintenance using in-
dustrial IoT.
Predictive maintenance bib-
liometric review.
Industrial IoT, pre-
dictive maintenance.
IoT improves industrial automation and asset
management.
[30]
Compare regional trends in IoT applica-
tions in education.
Bibliometric comparison of
IoT in education.
IoT in education, re-
gional trends.
IoT in education emphasizes regional dier-
ences in application.
[31]
Highlight evolving trends in IoT and Big
Data research.
Systematic review of biblio-
metric studies.
IoT, Big Data, the-
matic clusters.
IoT and Big Data have expanded across diverse
domains.
[32]
Analyze IoT in agriculture, emphasizing
food security.
Keyword and thematic
analysis of agriculture IoT.
IoT, agriculture, food
security.
IoT aids in food security and precision agricul-
ture.
[33]
Explore Industry 4.0 technologies in agri-
culture.
Keyword analysis and In-
dustry 4.0 mapping.
Industry 4.0, agricul-
ture, IoT.
Industry 4.0 enhances agricultural data man-
agement.
[34],
[35]
Study IoT in industrial automation and e-
ciency.
Cluster analysis and biblio-
metric review.
Industrial IoT, auto-
mation, maintenance.
IoT predictive maintenance reduces industrial
downtime.
Sensors 2025, 25, 906 15 of 38
[36]
Examine IoT’s impact on customer engage-
ment models.
Analysis of IoT-driven cus-
tomer models.
Customer engage-
ment, IoT, personali-
zation.
IoT drives customer personalization and secure
transactions.
[37]
Highlight IoT’s role in environmental sus-
tainability.
Bibliometric analysis of sus-
tainability applications.
Sustainability, IoT,
resource conserva-
tion.
IoT contributes to environmental monitoring
and energy conservation.
[38]
Focuses on anomaly detection in electricity
consumption using machine learning and
Big Data.
Machine learning algo-
rithms combined with Big
Data analysis.
Anomaly detection,
machine learning, Big
Data, electricity con-
sumption.
Identies anomalies in electricity consumption,
improving fraud detection and system e-
ciency using Big Data analytics and machine
learning.
[39]
Analyze IoT in healthcare focusing on real-
time monitoring.
Bibliometric study of
healthcare IoT.
Healthcare IoT, wear-
able technologies.
IoT enhances patient care but faces privacy and
interoperability issues.
[40],
[41]
Identify IoT and Big Data trends in sustain-
able agriculture.
Precision agriculture key-
word analysis.
Sustainability, agri-
culture, IoT.
IoT drives sustainable agriculture and resource
eciency.
[42]
Explore applications of WSNs and IoT in
Industry 4.0.
Systematic lliterature re-
view
Industry 4.0, IoT,
WSN.
IoT and WSN improve industrial eciency, en-
abling real-time monitoring and data-driven
decision making.
[43]
Analyze global trends in smart homes for
older adults.
Bibliometric and scien-
tometric analysis
Smart homes, older
adults, IoT, aging.
Technologies address health monitoring and
assistive living needs for aging populations.
[44]
Investigate smart technologies for sustaina-
ble urban water management.
Urban analysis
Smart technologies,
water management,
IoT.
IoT supports ecient water resource manage-
ment, promoting sustainability.
[45]
Examine IoT innovations for indoor air
quality monitoring.
Systematic review, biblio-
metric analysis
IoT, air quality, in-
door environment.
IoT enhances real-time environmental health
monitoring, beneting health outcomes.
[46]
Map a decade of research on smart homes
for elderly using scientometric methods.
Scientometric review,
CiteSpace
Smart homes, elderly,
research trends.
Insights highlight focus on fall detection,
health, and elderly comfort in smart home de-
signs.
[47]
Develop smart building strategies to en-
hance safety and health for the elderly.
Systematic and bibliometric
analysis
Smart buildings, el-
derly, safety, health.
Smart building designs address safety risks
and promote elderly health and well-being
[48]
Analyze IoT privacy and security chal-
lenges in smart homes.
Analytical review
IoT, privacy, security,
smart homes.
Privacy issues in IoT require robust security
measures to ensure user trust in smart home
systems.
[49]
Review smart home and city developments
concerning sustainability and future trends.
Comprehensive review
Smart homes, smart
cities, sustainability.
Integration of sustainable concepts in urban
and residential IoT systems is key for future
growth.
[50]
Systematically analyze trends and recom-
mendations for smart homes.
Systematic analysis
Smart homes, trends,
IoT.
Highlights future directions in smart home in-
novations, emphasizing user-centric and secure
designs.
[51]
To explore the conceptualization of smart
cities in the context of education, identify-
ing how education-related themes are inte-
grated into smart city frameworks.
Co-word analysis, thematic
clustering, keyword co-oc-
currence analysis
Smart cities, educa-
tion, urban develop-
ment, conceptualiza-
tion.
Education as a central component of smart city
research, oering thematic maps of the eld’s
evolution.
[52]
To explore the intellectual development of
smart city research using bibliometric and
main path analysis.
Bibliometric analysis, cita-
tion and co-citation analy-
sis, main path analysis
Smart cities, research
evolution, citation
networks, main path
analysis.
Evolution of smart city research, identifying
key works and research pathways. It empha-
sizes the intellectual ow of ideas within the
smart city domain.
[53]
To analyze the role of articial intelligence
(AI) in smart cities, focusing on trends and
relationships between key works in this
area.
Citation analysis, biblio-
graphic coupling, keyword
co-occurrence
AI, smart cities, ur-
ban management,
machine learning.
Integration of AI into smart cities, providing
insights into the technologies and applications
that are transforming urban management.
[54]
To provide a bibliometric analysis of smart
cities research, identifying key trends, re-
search clusters, and major contributors.
VOSviewer, CiteSpace, co-
authorship analysis, key-
word co-occurrence
Smart cities, co-au-
thorship, research
trends, interdiscipli-
nary collaboration.
Rapid growth in publications with emerging
themes like AI in governance.
[55]
To provide a bibliometric analysis of scien-
tic publications and patents on smart cit-
ies, bridging academic research with indus-
trial applications.
Bibliometric analysis, pa-
tent analysis, vantagepoint
Smart cities, patents,
scientic publica-
tions, industry appli-
cations.
Importance of the link between academic re-
search and industrial innovations.
The cross-analysis of the previous studies reveals several paerns and distinctions in
their objectives, methods, ndings and focus areas. Most studies share a common interest
in mapping research trends, highlighting interdisciplinary applications and identifying
research gaps. Studies such as [7,8,10] and [11] focus broadly on thematic mapping in IoT
and Big Data, whereas others, like [12,24,39] delve into specic sectors such as agriculture,
Sensors 2025, 25, 906 16 of 38
emphasizing precision farming, food security and sustainability. Healthcare is another
frequently explored area with studies like [25,27] focusing on wearable technologies and
real-time patient monitoring. Industrial IoT studies, including [29,34,35], emphasize pre-
dictive maintenance and operational eciency, demonstrating IoT’s impact beyond tra-
ditional IT applications. Ethical and technical challenges, such as privacy, data security
and system interoperability, are central to studies like [13–15].
Methodologically, bibliometric analysis dominates, as seen in [8,10,11] and others.
These studies often rely on tools like VOSviewer and SciMAT to map collaborations, topic
co-occurrences, and thematic clusters. Cluster analysis, used in [16,26], identies key
themes, particularly in smart city applications and infrastructure resilience. Some studies,
like [24], take a survey-based approach to provide broader overviews of Big Data and IoT
ecosystems.
The ndings across previous studies consistently highlight the interdisciplinary
growth of IoT and Big Data research. Studies like [7,9–11] observe their intersection with
various elds, including healthcare, agriculture and urban planning. Sector-specic in-
sights show that IoT signicantly enhances precision agriculture [12,28], patient monitor-
ing and healthcare eciency [27,39] and industrial automation [29,34]. Emerging trends,
such as sustainability-focused IoT applications [37,40,41] and smart city innovations, [26],
indicate growing interest in addressing global challenges like resource conservation and
urban management. However, challenges such as data scalability, integration and privacy
remain persistent, as highlighted in [13–15,25].
While many previous studies share similarities, such as recurring themes of real-time
analytics, collaboration networks and thematic exploration, they dier in their specic ar-
eas of focus. Some studies prioritize broad overviews [7,9], while others concentrate on
implementation in specic sectors [12,39]. Ethical and technical challenges are central to
studies like [13,14,25], whereas [34,35] emphasize operational benets and automation.
Several gaps remain across the reviewed studies. While challenges like interoperability
and scalability are widely recognized, actionable frameworks to address them are less ex-
plored. Emerging areas such as IoT in education [30] and blockchain’s integration with
IoT [25] are also underrepresented.
In terms of methodology, the last ve studies share a common reliance on biblio-
metric tools like CiteSpace, VOSviewer and Scopus, which are standard for mapping re-
search trends and visualizing networks of knowledge. However, the studies dier in the
scope of their focus and the tools they combine with bibliometric analysis. The authors of
[54] provide a comprehensive, longitudinal overview, while the authors of [51] focus on
the role of education within the smart city paradigm. Rejeb et al. introduce the novel use
of main path analysis to trace intellectual trajectories, adding depth to the bibliometric
approach [52]. Gupta et al. narrow their focus to AI applications, shedding light on a spe-
cic technological aspect of smart cities [53]. Puliga et al., on the other hand, incorporate
patent analysis, bridging the gap between academic research and industrial innovation,
oering a more comprehensive view of how smart city concepts are applied in practice
[55]. Each study contributes to the growing body of knowledge on smart cities, but they
do so using dierent methods that reect the diversity and complexity of the eld.
Hajoary et al. provide a broad mapping of the eld’s development, Scala et al. explore
specic educational frameworks, whereas Rejeb et al. use a methodological innovation to
map the ow of ideas through the research landscape [51,52,54]. Gupta et al. bring aen-
tion to the growing importance of AI in smart city development, while Puliga et al. oer
a dual analysis of academic publications and patents, presenting a more practical, indus-
try-oriented perspective [53,55].
To the best of the authors’ knowledge, the intersection of Big Data and IoT has not
yet been comprehensively examined through the combined methodologies of bibliometric
Sensors 2025, 25, 906 17 of 38
analysis and topic modeling. While previous studies (as presented in this section) have
separately explored advancements in Big Data and IoT and have employed bibliometric
tools to analyze research trends within these elds individually, a systematic and inte-
grated approach that investigates their intersection using advanced techniques such as
topic modeling remains unexplored. This gap is signicant because the convergence of
these two domains represents a critical area of technological innovation, inuencing di-
verse applications such as smart cities, healthcare and industrial automation.
By employing bibliometric analytics alongside topic modeling, such as Latent Di-
richlet Allocation (LDA), it becomes possible to uncover not only the dominant research
themes and trends but also the evolving relationships between these interconnected elds.
How Big Data and IoT research areas complement each other is revealed, identifying the-
matic clusters at their convergence and highlighting emerging research trajectories that
traditional bibliometric analyses might overlook.
3. Research Methodology
The dataset used in our research encompasses nine full record les, collectively con-
taining information on 8159 publications related to Big Data and IoT research. These rec-
ords have been sourced from Web of Science, a leading research database that includes
high-quality, peer-reviewed publications. The timeframe covered by the dataset captures
the evolution of research in these elds, providing a broad overview of key developments
and trends over the years.
Each data le provides essential bibliographic details such as author names, article
titles and information about the conference or journal where the research was presented
or published. This allows for a complex analysis of collaboration paerns among research-
ers and institutions as well as an understanding of the most inuential venues for dissem-
inating knowledge in Big Data and IoT.
Additionally, the datasets contain keywords and abstracts that reveal the main
themes and highlights about all areas of the research. These textual elements are particu-
larly valuable for NLP techniques, which will be used to analyze paerns and trends in
the text of the literature. By extracting key terms and recurring topics, it becomes possible
to identify emerging research areas and map how certain themes have evolved.
Overall, this dataset provides a foundation for exploring the collaborative networks
between authors and institutions, uncovering key trends within the literature and con-
ducting an analysis of the content through NLP. By leveraging this dataset, our research
aims to oer a detailed overview of the academic landscape in Big Data and IoT, contrib-
uting to a clearer understanding of how research in these elds is developing and where
future opportunities for innovation may lie.
The research methodology is described in Figure 2. First, data are collected from the
Web of Science platform, comprising 8159 publications from nine full record les within
the time interval of 2010 to 2024. The results were obtained using the following search:
* Internet of Things Arcle English
Retracted Publicaon
where
AF—All Fields;
DT—Document Types;
OA—Open Access;
LA—Language;
RN—Retraction Notices;
PY—Publication Years;
Sensors 2025, 25, 906 18 of 38
*—any word following big data;
˄—AND;
¬—NOT.
Thus, the targeted publications are at the intersection of these two concepts: Big Data
and IoT. Publications that focused only on Big Data or IoT were not included in the input
data. Bibliographic data include essential elds such as author names, article titles, con-
ference or journal information, and publication dates. Next, the Data Processing stage in-
volves cleaning the data through steps like missing value removal, stop words removal,
tokenization, and lemmatization, employing tools such as NLTK, Gensim and the Word-
Net Lemmatizer. In the Sentiment Analysis phase, sentiment trends over time are evalu-
ated using custom keyword matching, and a bar chart is generated to visualize the senti-
ment distribution.
An alternative strategy for sentiment analysis in specialized elds is to create custom
word lists tailored to the specic context of the research. This process involves selecting
and categorizing words and phrases that are directly relevant to topics such as Big Data
and IoT. Sentiment scores are then assigned to these terms based on their contextual im-
plications. For example, in this domain, words like “innovation”, “performance”, “e-
ciency”, or “breakthrough” could signify positive sentiment, while terms such as “uncer-
tainty”, “decline”, or “challenge” might indicate negative sentiment. Neutral terms might
consist of technical or descriptive language that carries no inherent emotional weight. This
approach is advantageous for exploring specialized or niche areas, where generic senti-
ment analysis tools may overlook critical nuances in terminology.
For Topic Modeling, the LDA technique is used to extract topics along with
WordCloud for keyword visualization. Hyperparameters, such as the number of topics
and coherence score, are tuned using tools like Gensim, WordCloud, and pyLDAvis.
Finally, the Results and Interpretations stage identies collaboration networks,
emerging research themes, and provides opportunities for future research.
Figure 2. Methodology owchart.
Sensors 2025, 25, 906 19 of 38
In our research, NLP techniques are applied to analyze the large volume of textual
data related to Big Data and IoT research. Given the vast number of abstracts, titles and
keywords, these methods help extract key concepts, identify paerns and uncover im-
portant trends within literature.
The rst step in the data processing involves breaking down the text into smaller
units using tokenization, allowing for the identication of frequently mentioned terms.
To ensure consistency and reduce redundancies, processes like stemming and lemmatiza-
tion are applied, which normalize words by reducing them to their base forms.
A major part of the analysis involves topic modeling, which is a technique that helps
detect underlying themes across the body of literature. By doing so, it will be possible to
identify the primary areas of research within Big Data and IoT as well as track how these
topics have evolved. Additionally, our research employs techniques like TF-IDF to high-
light the most relevant and distinctive terms within the dataset, providing insights into
specialized areas of research.
The data processing and analysis are carried out using Python, along with specialized
libraries for text analysis and data manipulation, ensuring an ecient and structured ex-
ploration of the dataset. Together, these methods oer an understanding of the research
landscape in Big Data and IoT. The methodological steps and parameters are summarized
in Table 2.
Table 2. Methodology ow using parameters.
Step
Description
Details
1
Seing Objective
Analyze the main themes and sentiment across a dataset of IoT and
Big Data publications.
Time Frame: Various years of publication. Data
Source: Web of Science
2
Data Collection
Collected bibliographic data from Web of Science.
Total Records: 8159 from nine distinct datasets.
3
Data Preparation
Preprocessing of text data for analysis.
Text cleaning, stop word removal, tokenization,
and lemmatization.
4
Sentiment Analysis
Assessed sentiment of each document based on abstract content.
Sentiment categories: Positive, Negative, Neu-
tral. Sentiment distribution visualized.
5
Topic Modeling
Applied LDA to uncover latent themes within the dataset.
Generated six topics, rened with coherence
scoring and visualized.
6
Visualization
Displayed sentiment and topic distribution using graphs and word
clouds.
Tools: WordCloud for keyword analysis, LDA
visualization with pyLDAvis.
7
Interpretation and
Conclusion
Provided insights on prominent research themes and sentiment trends.
Highlighted dominant topics and emerging re-
search areas.
This structured approach presents the methodology ow, detailing each step of the
analysis with specic parameters. The importance of integrating validation steps, even in
fully automated studies, is obvious, in order to ensure that conclusions drawn from large
datasets are robust and defensible. Studies like Systematic Mapping Studies (SMSs) or
similar approaches that summarize the existing literature are reliable. While our research
relies on automated NLP techniques with a human-validation step, the proposed meth-
odology has both advantages and limitations compared to manual or hybrid methods
commonly used in SMS. Automated vs. manual processes pose interesting challenges. The
exclusive reliance on NLP techniques ensures objectivity and scalability, allowing for the
analysis of large datasets like the 8159 publications. Automation minimizes biases intro-
duced by human decision making, particularly during source inclusion/exclusion. How-
ever, as noted, the absence of manual screening and verication can raise concerns about
the reliability of the conclusions. Manual checks by multiple researchers, as seen in tradi-
tional SMSs, provide a form of triangulation that strengthens the validity of the results.
To enhance the credibility of an automated NLP-driven approach, we have included a
validation step to cross-check results. For instance, we compared a subset of NLP-identi-
ed themes with manual interpretations by domain experts, ensuring alignment between
automated insights and human understanding. While the study leverages the power of
Sensors 2025, 25, 906 20 of 38
NLP to process a vast dataset eciently, integrating a hybrid approach (automated pro-
cessing supplemented by selective manual validation) strengthens its reliability. For ex-
ample, we manually reviewed a small percentage of abstracts or themes to assess the ac-
curacy of the NLP-generated outputs.
4. Results
The analysis of publication trends in Big Data and IoT presented in Figure 3 reveals
a landscape dominated by original research with 54.7% of contributions categorized as
Articles. This indicates that researchers are actively generating new knowledge and driv-
ing innovation in these evolving elds. Additionally, 35.9% of the publications are Pro-
ceedings Papers, which highlight the essential role of conferences in facilitating
knowledge exchange and collaboration among professionals.
The representation of Review articles, at 8.7%, indicates a potential opportunity for
researchers to undertake systematic reviews or meta-analyses that synthesize existing
knowledge and provide comprehensive overviews to guide future inquiries in Big Data
and IoT research. Furthermore, the minimal 0.7% in the “Others” category emphasizes a
clear focus within the research community on traditional publication formats.
Overall, these ndings illustrate a dynamic research environment, showcasing the
importance of original contributions while also highlighting the need for more synthe-
sized knowledge to support ongoing and future inquiries in Big Data and IoT.
Figure 3. Distribution of document types.
In Figure 4, the number of publications from each year can be observed. It illustrates
paerns in research output over the years, with a notable peak in 2022, indicating a surge
in interest and innovation in Big Data and IoT. This increase may be linked to technolog-
ical advancements or emerging trends. The subsequent decline from 2022 to 2024 raises
questions about inuencing factors, while the steady growth observed from 2013 to 2020
reects increasing investment and an expanding research community.
Sensors 2025, 25, 906 21 of 38
Figure 4. Number of publications per year.
Figure 5 highlights a disparity in the number of articles published across dierent
journals, showing that over 3000 articles have been published in IEEE journals. In com-
parison, Elsevier, Springer, and MDPI each account for between 500 and 1000 articles,
while other publishers contribute fewer than 500 articles. This distribution is signicant
for the article as it underscores the dominant role of IEEE in the elds of technology and
engineering research.
Figure 5. Number of articles per publisher.
Figure 6 illustrates the evolution of publications in the top journals publishing on IoT
and Big Data topics from 2016 to 2024. The data reveal minimal contributions across jour-
nals prior to 2016, which is followed by a signicant rise in publications. Between 2018
and 2020, publication numbers peaked with journals like IEEE Internet of Things Journal
and Sensors playing prominent roles. Similarly, IEEE Access and Wireless Communications
experienced substantial growth. Post-2020, the trends stabilize, with journals such as Ap-
plied Sciences-Basel and Sustainability maintaining signicant contributions. This trend in-
dicates a growing interest in elds related to wireless communications, sustainability, and
IoT over recent years.
Sensors 2025, 25, 906 22 of 38
Figure 6. Top 10 publication trends by journals.
Figure 7 showcases WOS categories network connections. Each node represents a
specic research category, and the size of the nodes reects the degree of co-occurrence of
the respective category with others. The connections or edges between the nodes illustrate
how often pairs of categories appear together within the same literature.
For instance, categories like “Telecommunications” and “Engineering, Electrical &
Electronic” are prominently connected, indicating an overlap in research between these
areas. Other closely linked elds, such as “Computer Science, Information Systems” and
“Computer Science, Theory & Methods”, show similar co-occurrence paerns. The visu-
alization highlights key interdisciplinary research clusters, particularly around electrical
engineering and computer science topics, where the elds most commonly intersect.
This interconnectedness underscores the signicance of interdisciplinary collabora-
tions in driving IoT and Big Data research forward. Our ndings emphasize that such
collaborations are signicant for addressing complex challenges and fostering innovation.
Fields such as computer science and engineering, which dominate the co-occurrence net-
work, converge to develop technologies for real-time data processing and scalable IoT
frameworks, which are capabilities essential for applications like industrial IoT and smart
cities. These collaborations enable advancements in system eciency, data integration
and scalability, addressing the practical demands of these domains. By bridging expertise
across categories, as illustrated in Figure 7, the research highlights how interdisciplinary
research strengthens the foundation for transformative solutions in IoT and Big Data.
Sensors 2025, 25, 906 23 of 38
Figure 7. WoS categories network connections.
The word cloud presented in Figure 8 highlights the most frequently occurring terms
in the analyzed publications with “Internet”, “IoT Thing”, “Big Data”, “network” and
“computing” standing out as the dominant keywords. These terms reect the core areas
of focus within the research landscape of Big Data and IoT, indicating the primary tech-
nological components and concepts that drive innovation in these elds. For instance, “In-
ternet” and “IoT Thing” emphasize the importance of connectivity and the integration of
physical devices in the Internet of Things, while “Big Data” and “computing” underscore
the role of large-scale data processing and computational power in supporting these sys-
tems.
Figure 8. Word cloud graphic representation of the main keywords.
The sentiment analysis of abstracts, as shown in the bar chart in Figure 9, reveals a
clear dominance of positive sentiment within the dataset. Over 7000 abstracts are classi-
ed as positive, while fewer than 1000 are identied as negative. Neutral abstracts consti-
tute a negligible fraction in comparison. This distribution aligns with the typical purpose
of abstracts, which is to emphasize the signicance, advancements, and optimistic out-
comes of the research, reecting a generally positive tone in Big Data and IoT studies.
Sensors 2025, 25, 906 24 of 38
Positive abstracts focus on showcasing innovations, solutions to existing challenges or
promising results that highlight the contributions of the research.
Negative sentiments, on the other hand, primarily arise from discussions of unre-
solved issues or barriers in the eld. These include challenges such as the limited scalabil-
ity of existing frameworks, the lack of robust interoperability solutions, and ongoing dif-
culties in eciently processing large volumes of real-time data. Such abstracts often em-
phasize the limitations or risks that impede the broader adoption or application of IoT
technologies.
Neutral sentiments are linked to more descriptive or method-focused studies that
neither highlight advancements nor identify major challenges. These abstracts frequently
appear in conference publications, which may lack in-depth analysis or the practical vali-
dation of proposed solutions. Additionally, gaps in systematic review studies lead to frag-
mented knowledge, making it dicult to synthesize a comprehensive understanding of
progress in the eld.
Figure 9. Sentiment analysis of abstracts.
Figure 10 displays the average distribution of six topics extracted from the Abstract
column. Topic 1 has the highest average distribution, approximately 20%, suggesting that
themes related to “smart systems”, “IoT”, and “data” are the most frequently discussed
in the dataset. Topic 4 follows closely, with an average distribution of around 18%, high-
lighting a focus on areas like “IoT”, “cloud computing”, and “edge computing”. Topics 5
and 6, with average distributions of about 17% and 16% respectively, indicate aention
toward “security models”, “learning systems”, and “industrial IoT applications.” On the
other hand, Topic 2 has the lowest average distribution, around 13%, suggesting that
themes related to “data processing”, “sensors”, and “proposed technologies” are less
prominent in this dataset. Topic 3, with an average distribution of 15%, reects research
areas such as “Big Data”, “healthcare applications”, and “IoT in healthcare”. Therefore,
Figure 10 illustrates how the identied topics are proportionally distributed across the
dataset. The average distributions indicate the extent to which each topic is covered in the
analyzed documents. A higher average distribution, as seen in Topic 1, suggests that the
themes associated with this topic are more frequently discussed and appear in more doc-
uments. Conversely, lower values, such as those for Topic 2, indicate that these topics are
less frequently present in the dataset.
Sensors 2025, 25, 906 25 of 38
Figure 10. Average topic distribution.
The six identied topics overlap in several areas (as in Table 3), reecting shared
themes and dependencies across research areas. For instance, Topic 1 focuses on “smart
systems” and “IoT”, which require the infrastructure addressed in Topic 4, such as “cloud
computing” and “edge computing”, for data management and processing. Topic 3, cen-
tered on “healthcare applications”, is closely related to Topic 6, which addresses “learning
systems” that are essential for analyzing healthcare data and enabling predictive solu-
tions. Topics 4 and 5 are also linked, as “security models” from Topic 5 are critical for
securing cloud and edge computing frameworks described in Topic 4. Although Topic 2
has the smallest distribution, its focus on “data processing” and “sensors” serves as a
foundational element that supports the systems described in the other topics.
Table 3. Overview of topics, top words and average distributions.
Topic
Top Words
Average Distribution
1
smart, IoT, data, internet, systems
0.2001862
2
data, IoT, sensor, proposed, processing
0.12894942
3
data, big, IoT, healthcare, internet
0.15653505
4
IoT, computing, data, cloud, network, edge
0.1803717
5
data, IoT, security, model, proposed
0.17459249
6
industry, technologies, research, IoT
0.15936514
The perplexity plot, presented in Figure 11, helps evaluate the quality of the LDA
model by ploing the perplexity values across dierent numbers of topics. As shown in
the graph, the perplexity decreases consistently as the number of topics increases from
two to six with the lowest perplexity achieved at six topics. This indicates that six topics
provide the best t for this dataset. The steady decline in perplexity suggests that increas-
ing the number of topics improves the model’s ability to capture the structure of the da-
taset without signs of overing.
Sensors 2025, 25, 906 26 of 38
Figure 11. Perplexity vs. number of topics.
In Figure 12, the bar chart illustrates the most important words within Topic 1. Words
like “data”, “IoT”, “proposed” and “network” are prominent, reecting the central themes
of IoT and Big Data in this topic.
.
Figure 12. Topic 1 word distribution.
The heatmap presented in Figure 13 visualizes the alpha and beta parameters across
six topics, providing insights into their distribution and signicance. The alpha parameter
controls the document–topic distribution, indicating how frequently a topic appears
across the dataset. Higher alpha values suggest that a larger number of documents are
likely to include these topics, highlighting their relevance within the dataset. Topics 1, 3
and 6 exhibit relatively higher alpha values, suggesting their signicance and broader
coverage across the collection of dataset analyzed. The beta parameter, on the other hand,
inuences the topic–word distribution, reecting the diversity of words associated with
each topic. A higher beta value signies that a topic encompasses a broader and more
Sensors 2025, 25, 906 27 of 38
diverse vocabulary, making it less narrowly focused. Among all topics, Topic 5 has the
highest beta value, indicating that it includes the most varied range of words. This makes
Topic 5 more diverse in scope compared to the other topics, potentially representing a
broader or more general theme within the dataset.
The visual contrast in the heatmap, represented by the intensity of the color gradient,
emphasizes these dierences. For instance, the darker blue shading for Topic 5 in the beta
row highlights its signicantly higher value, underlining its diversity. In contrast, the
lighter shades for alpha in Topics 2 and 4 demonstrate their relatively lower values, indi-
cating that these topics are less prevalent across the dataset.
Figure 13. Alpha and beta parameters.
Table 4 reects the distribution of six main topics extracted from an LDA model ap-
plied to the analyzed dataset. Each topic is associated with three key values: the average
distribution of the topic across the entire dataset, the alpha parameter, and the beta pa-
rameter. The average distribution shows how frequently each topic appears in the ana-
lyzed texts. In this case, Topic 3 and Topic 1 have the highest distributions, both around
0.246, indicating that these topics are predominant in the dataset. In contrast, Topic 2 has
the lowest distribution at 0.1438, suggesting that this topic is less frequently discussed in
the abstracts.
Table 4. Alpha and beta impact on topic distribution stability.
Topic
Average Distribution
Alpha
Beta
Topic 1
0.2462
0.1
0.15
Topic 2
0.1438
0.05
0.12
Topic 3
0.2466
0.1
0.2
Topic 4
0.1853
0.08
0.18
Topic 5
0.178
0.07
0.25
Topic 6
0.2142
0.9
0.22
The alpha parameter controls how spread out the topics are within each document.
Higher alpha values, such as those for Topic 1 and Topic 3 (both with a value of 0.10),
indicate that dataset contains a more balanced proportion of multiple topics. On the other
Sensors 2025, 25, 906 28 of 38
hand, Topic 2, with an alpha value of 0.05, appears in fewer datasets and tends to be pre-
sent in smaller proportions within the texts where it is included.
The beta parameter refers to the distribution of words within each topic. A higher
beta suggests that the respective topic is composed of a wider range of words. For exam-
ple, Topic 5, with a beta of 0.25, includes a broader variety of terms compared to Topic 1,
which has a lower beta of 0.15, indicating a more limited and focused vocabulary centered
around specic terms.
In Figure 14, a visual representation of topics derived from the abstracts of research
papers is presented. Using LDA, the content of these abstracts was grouped into six dis-
tinct topics, and t-SNE (t-distributed Stochastic Neighbor Embedding) was applied to pro-
ject these high-dimensional topic distributions into a two-dimensional space. This makes
it easier to visually understand the relationships and distinctions between the topics.
In the graph, each region corresponds to a dierent topic with transitions between
colors showing how some documents may overlap across multiple themes. For example,
articles that discuss both the Internet of Things (IoT) and Big Data might lie in areas where
the colors blend, while articles focusing solely on one of these subjects are placed in more
distinct regions.
Figure 14. A 2D t-SNE projection of topics.
From Figure 15, a steady rise in coherence from two topics (0.32) to a peak at ve
topics (0.44) is noticed, suggesting that the topics at this point are the most interpretable
and meaningful. After the peak, the coherence score drops at six topics (0.36), followed by
uctuations, reaching another high point at nine topics (0.41).
Sensors 2025, 25, 906 29 of 38
.
Figure 15. Coherence according to the number of topics.
Figure 16 illustrates the compound annual growth rate (CAGR) for the top keywords,
showing how their usage has evolved over time. CAGR represents the rate at which a
value grows annually over a specic period. In this case, the keyword “IoT” shows the
highest growth, indicating that it has become increasingly common in the analyzed da-
taset. The terms “internet” and “thing” follow, also reecting signicant increases in us-
age. Keywords like “big”, “data”, “system” and “network” have grown at a steady pace,
while “device”, “technology” and “smart” exhibit lower growth rates compared to the
others.
Figure 16. Abstract keywords evolution over time.
Figure 17 illustrates the trends of Big Data, IoT, AI, Cloud and Edge. The x-axis de-
notes the years, while the y-axis represents the count of mentions, providing an intuitive
visualization of how each topic’s popularity has uctuated over time. Each year is repre-
sented as a group of bars with each bar in the group corresponding to one of the ve
research topics. Certain topics such as “AI” and “IoT” show consistent activity with peaks
in specic years that suggest heightened research focus or breakthroughs. Topics like
“Edge” and “Cloud” demonstrate periodic surges, indicating their growing importance
Sensors 2025, 25, 906 30 of 38
in recent years. On the other hand, “Big Data” maintains a relatively stable trend, high-
lighting its sustained relevance throughout the period.
Figure 17. Number of mentions over time.
Table 5 presents the top terms and their respective weights across six dierent topics.
Each row represents a specic topic, and within each row, the “Score” columns combine
the key terms with their associated weights, providing an understanding of the im-
portance of each term within the topic. For instance, Topic 0 is characterized by terms like
“data” (0.011), “IoT” (0.009), and “technologies” (0.009), suggesting a focus on technolo-
gies and IoT. Similarly, Topic 1 highlights “smart” (0.012) and “data” (0.009) as central
concepts, indicating discussions around smart technologies and their data integration.
Across all topics, key themes related to data, IoT, and technologies are notable, reecting
their signicance in the broader discourse. The weights aached to each term reect the
strength of their association within each topic, allowing for a more detailed understanding
of the underlying thematic structure (as in Table 5).
Table 5. Key terms and their weights across topics.
Topic
No.
Score 1
Score 2
Score 3
Score 4
Score 5
Score 6
Score 7
Score 8
Score 9
Score 10
0
data
(0.011)
IoT
(0.009)
technologies
(0.009)
industry
(0.008)
research
(0.007)
smart
(0.006)
digital
(0.006)
systems
(0.006)
paper
(0.006)
system
(0.005)
1
smart
(0.012)
data
(0.009)
IoT
(0.007)
system
(0.007)
paper
(0.006)
model
(0.005)
technology
(0.005)
research
(0.005)
industry
(0.005)
information
(0.004)
2
research
(0.013)
industry
(0.012)
technologies
(0.010)
data
(0.010)
study
(0.006)
technology
(0.006)
IoT
(0.005)
computing
(0.005)
supply
(0.005)
paper
(0.005)
3
data
(0.023)
IoT
(0.010)
system
(0.007)
energy
(0.007)
smart
(0.007)
proposed
(0.006)
paper
(0.006)
big
(0.006)
network
(0.005)
devices
(0.005)
4
data
(0.022)
IoT
(0.009)
learning
(0.007)
proposed
(0.006)
model
(0.005)
results
(0.005)
things
(0.005)
internet
(0.005)
devices
(0.004)
based
(0.004)
5
data
(0.015)
model
(0.013)
proposed
(0.008)
system
(0.008)
using
(0.007)
energy
(0.006)
based
(0.006)
ber
(0.006)
results
(0.006)
algorithm
(0.005)
In the following, the six topics generated from the dataset will be analyzed in detail,
using the pyLDAvis tool for topic modeling visualization. pyLDAvis is an interactive tool
that provides a clear, visual representation of topics created using LDA. The tool helps
explore relationships between the topics and highlights the most relevant terms for each.
This complements the earlier table, “Key Terms and Their Weights Across Topics,” by
Sensors 2025, 25, 906 31 of 38
oering an interactive and detailed view of the term distribution and the thematic struc-
ture within the dataset.
LDA, a widely used topic modelling algorithm, was applied to extract latent topics
from the dataset. This probabilistic approach identies groups of co-occurring terms that
frequently appear together, revealing distinct themes embedded in the data. The number
of topics (6) was determined through iterative experimentation. Topic coherence, a metric
used to measure the semantic consistency of topics, was a key determinant in selecting the
optimal number of topics. A higher coherence score (0.79) suggests that the terms within
each topic are meaningfully related, enhancing the quality of the topics.
Figure 18 focuses on Topic 1, the most dominant one, as shown by the larger red
circle on the left side of the visualization. The intertopic distance map (on the left) shows
the proximity of dierent topics with larger circles representing the most common topics
in the dataset. Topic 1 stands out as a signicant theme, indicated by its larger circle and
central position, suggesting it covers a broad range of discussions in the dataset.
On the right side, the top 30 most relevant terms for Topic 1 are displayed. Words
like “data”, “IoT”, “system” and “energy” are among the most frequent, signaling that
Topic 1 revolves around data systems and IoT technologies, likely in the context of smart
systems and energy-related applications. The bar chart highlights the relevance of these
terms within the topic (red portion) and compares their overall frequency in the entire
corpus (blue portion).
Figure 18. PyLDAvis: the rst topic.
Topic 2, presented in Figure 19, focuses on the application of data and IoT technolo-
gies, as indicated by terms such as “technologies”, “industry” and “research”. This topic
deals with how IoT and related technologies are transforming various industries, includ-
ing manufacturing and energy sectors. Terms like “manufacturing” and “energy” suggest
that industrial processes and energy management are key areas of focus within this topic.
The presence of terms like “systems”, “model” and “applications” hints at the technical
and applied nature of the research, pointing toward discussions around the development
and deployment of models and systems in practical seings.
Sensors 2025, 25, 906 32 of 38
Figure 19. PyLDAvis: the second topic.
The third topic (in Figure 20) emphasizes terms like data, IoT, and learning, indicat-
ing a focus on machine learning and IoT applications. The prominence of terms such as
“proposed”, “model” and “results” suggests that this topic is oriented toward research
involving new methods and models in the IoT domain, which is related to learning algo-
rithms or clustering techniques. The appearance of terms like “clustering”, “algorithm”
and “detection” highlights the technical focus on processing and analyzing data within
IoT systems.
Figure 20. PyLDAvis: the third topic.
The fourth topic (in Figure 21) highlights terms such as smart, data, and IoT, with
“smart” being the main term, suggesting a focus on smart technologies and systems. The
presence of terms like “system”, “model” and “network” points to discussions surround-
ing smart systems, their applications, and management. Additionally, terms like “tech-
nology” and “research” reect an ongoing focus on technological advancements and their
implications in various industries, including manufacturing and network systems.
Sensors 2025, 25, 906 33 of 38
.
Figure 21. PyLDAvis: the fourth topic.
Topic 5, presented in Figure 22, highlights the intersection of research and industry,
focusing on technologies, data, and innovation. It also touches upon the role of digital
chains and supply systems, suggesting an industrial focus on digital transformation. The
presence of terms like “applications”, “analysis” and “innovation” underscores the prac-
tical implications of these technologies within industry seings.
.
Figure 22. PyLDAvis: the fth topic.
Figure 23 presents the last topic, Topic 6, which focuses on technical aspects such as
data modeling, system performance, and prediction algorithms. It delves into the e-
ciency of IoT networks with terms like “accuracy”, “power” and “performance” standing
out, indicating the aention to optimization in IoT infrastructures. Additionally, this topic
emphasizes the use of algorithms to enhance system control and accuracy, showcasing
the depth in IoT performance improvements.
Sensors 2025, 25, 906 34 of 38
Figure 23. PyLDAvis: the sixth topic.
Figures 18–23 also illustrate how these topics blend in research that spans multiple
domains with transitions between themes indicating overlap in areas such as network ef-
ciency, real-time analytics, and scalable architectures. This interconnectedness under-
scores how advancements in one topic, such as machine learning techniques from Topic
3, can enhance the predictive capabilities discussed in Topic 6 or the operational eciency
highlighted in Topics 2 and 5.
5. Conclusions
Bearing in mind the goals formulated in the introduction, we extracted the following
main ndings:
• Trends over the years indicate a marked increase in publication output, peaking in
2022. This surge is likely linked to rapid technological advancements and heightened
interest in these elds. Nevertheless, the decline observed from 2022 to 2024 prompts
further investigation into the factors inuencing research momentum, suggesting a
potential plateau in interest or resource allocation.
• The analysis of publications by source reveals a striking concentration within IEEE
journals, which dominate the eld with over 3000 articles published. This dominance
underscores IEEE’s signicant role in advancing technology and engineering re-
search. In contrast, other publishers, including Elsevier, Springer and MDPI, contrib-
ute substantially fewer articles, indicating a potentially skewed landscape that favors
specic publication venues.
• The visualization of keyword co-occurrences and sentiment analysis demonstrates
the interdisciplinary nature of research in Big Data and IoT with strong links between
telecommunications, engineering and computer science. The prevalent positive sen-
timent across the abstracts suggests that the prevailing research narrative is optimis-
tic, highlighting advancements and benecial trends within these elds.
• The topic modeling using LDA indicates six prominent themes, including smart sys-
tems, industrial applications and machine learning, illustrating the multifaceted fo-
cus of current research. The evolving usage of key terms over time, particularly the
rise of “IoT”, reects the growing integration of IoT concepts into various technolog-
ical domains. The PyLDAvis visualizations analyze six main topics within a dataset
each with unique thematic focuses. The rst topic is the most dominant, centered on
data systems and IoT technologies, particularly in smart systems and energy
Sensors 2025, 25, 906 35 of 38
applications, with frequent terms like “data”, “IoT”, “system” and “energy.” The sec-
ond topic explores IoT applications in various industries, specially manufacturing
and energy, highlighting terms like “technologies”, “industry” and “research”. The
third topic emphasizes IoT and machine learning, focusing on new methods and
models involving algorithms and clustering techniques. The fourth topic centers on
smart technologies, discussing systems, models, and networks, with terms such as
“smart” and “system”. The fth topic examines the intersection of research, technol-
ogy, and industry, focusing on digital transformation in supply systems and innova-
tion. The sixth topic delves into technical IoT aspects like system performance and
prediction algorithms with an emphasis on optimizing IoT infrastructure for accu-
racy and control. Each topic’s terms and themes illustrate dierent facets of IoT and
data technologies, ranging from technical implementations to industry applications.
LDA has several limitations that can inuence the reliability and interpretability of
our analysis on Big Data and IoT research. One of the main challenges is its sensitivity to
hyperparameters, such as the number of topics (K). The choice of K signicantly aects
the coherence and distinctiveness of topics. In this research, the selection of six topics was
optimized tuning the hyperparameters of the LDA model in order to identify the best K.
Another issue with LDA is its reliance on probabilistic word distributions rather than deep
semantic understanding. The model does not dierentiate between synonymous terms or
recognize broader conceptual relationships. Consequently, similar concepts, such as “pre-
dictive maintenance” and “fault detection”, might be assigned to separate topics despite
their close thematic connection. Despite these limitations, LDA is a valuable tool for ex-
tracting overarching themes in research. However, to enhance the robustness of the anal-
ysis, our future research will explore alternative approaches, such as BERTopic or neural
topic models like BERT-LDA to oer beer semantic understanding.
Our current research limited its dataset to Web of Science publications due to the
platform’s well-established reputation for indexing high-quality, peer-reviewed literature
and its comprehensive coverage across a wide range of disciplines.
In conclusion, while the landscape of Big Data and IoT research is marked by dy-
namic innovation and signicant original contributions, it also presents notable gaps in
synthesized knowledge that warrant aention. Future researchers should focus on devel-
oping comprehensive review articles that could bridge these gaps, guiding subsequent
inquiries. The ongoing evolution of publication types and trends emphasizes the need for
adaptability in research dissemination strategies, ensuring that diverse formats are uti-
lized to eectively communicate ndings to a broader audience.
Author Contributions: All authors contributed to the study conception and design. Material prep-
aration, data collection and analysis were performed by D.-A.C. and S.-V.O. Literature review and
supervision were ensured by A.-A.C. and A.B. The rst draft of the manuscript was wrien by all
authors, and they also commented on previous versions of the manuscript. All authors read and
approved the nal manuscript.
Funding: This work was supported by a grant of the Ministry of Research, Innovation and Digitiza-
tion, CNCS/CCCDI–UEFISCDI, project number COFUND-CETP-SMART-LEM-1, within PNCDI IV.
Acknowledgements: This work was supported by a grant of the Ministry of Research, Innovation
and Digitization, CNCS/CCCDI–UEFISCDI, project number COFUND-CETP-SMART-LEM-1,
within PNCDI IV.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data will be made available upon request.
Sensors 2025, 25, 906 36 of 38
Conicts of Interest: The authors declare no conict of interest.
References
1. Fawzy, D.; Moussa, S.M.; Badr, N.L. The Internet of Things and Architectures of Big Data Analytics: Challenges of Intersection
at Different Domains. IEEE Access 2022, 10, 4969–4992.
2. Elgendy, N.; Elragal, A. Big Data Analytics: A Literature Review Paper. In Advances in Data Mining; Applications and Theoretical
Aspects; Springer: Berlin/Heidelberg, Germany, 2014.
3. Asatani, K.; Mori, J.; Ochi, M.; Sakata, I. Detecting trends in academic research from a citation network using network represen-
tation learning. PLoS ONE 2018, 13, e0197260.
4. Bhuvana, P.; Nandhini, J.; Gnanasekaran, T. An Analysis of the Applications of Natural Language Processing in Various Sectors.
In Smart Intelligent Computing and Communication Technology; IOS Press: Amsterdam, The Netherlands, 2021.
5. Sharma, R.; Agarwal, P.; Arya, A. Natural Language Processing and Big Data: A Strapping Combination. In New Trends and
Applications in Internet of Things (IoT) and Big Data Analytics; Springer: Berlin/Heidelberg, Germany, 2022; pp. 255–271.
6. Amin, F.; Abbasi, R.; Khan, S.; Abid, M.A.; Mateen, A.; de la Torre, I.; Castilla, A.K.; Villena, E.G. Latest advancements and
prospects in the next-generation of Internet of Things technologies. PeerJ Comput. Sci. 2024, 10, e2434.
7. Chataut, R.; Phoummalayvane, A.; Akl, R. Unleashing the Power of IoT: A Comprehensive Review of IoT Applications and
Future Prospects in Healthcare, Agriculture, Smart Homes, Smart Cities, and Industry 4.0. Sensors 2023, 23, 7194.
8. Ö zköse, H. Bibliometric Analysis and Scientific Mapping of IoT. J. Comput. Inf. Syst. 2023, 63, 1438–1459.
9. Oyewola, D.O.; Dada, E.G. Exploring machine learning: A scientometrics approach using bibliometrix and VOSviewer. SN Appl.
Sci. 2022, 4, 143.
10. Goranin, N.; Hora, S.K.; Čenys, H.A. A Bibliometric Review of Intrusion Detection Research in IoT: Evolution, Collaboration,
and Emerging Trends. Electronics 2024, 13, 3210.
11. Parlina; Ramli, K.; Murfi, H. Theme Mapping and Bibliometrics Analysis of One Decade of Big Data Research in the Scopus
Database. Information 2020, 11, 69.
12. Abdullahi, H.; Mahmud, M.; Hassan, A.; Ali, A. A bibliometric analysis of the evolution of IoT applications in smart agriculture.
Ingénierie Des Systèmes D’information 2023, 28, 1495–1504.
13. Mehdipour, F. A Review of IoT Security Challenges and Solutions. In Proceedings of the 8th International Japan-Africa Confer-
ence on Electronics, Communications, and Computations (JAC-ECC), Alexandria, Egypt, 18–20 December 2020.
14. Virat, M.; Bindu, S.; Aishwarya, B.; Dhanush, B.; Kounte, M. Security and Privacy Challenges in Internet of Things. In Proceed-
ings of the 2018 2nd International Conference on Trends in Electronics and Informatics (ICOEI), Tirunelveli, India, 11–12 May
2018.
15. Aljarrah, M.; Zawaideh, F.; Magableh, M.; Wahshat, H.A.; Mohamed, R.R. Internet of Thing (IoT) and Data Analytics with
Challenges and Future Applications. In Proceedings of the 2023 International Conference on Computer Science and Emerging
Technologies (CSET), Bangalore, India. 10–12 October 2023.
16. Alaeddini, M.; Hajizadeh, M.; Reaidy, P. A Bibliometric Analysis of Research on the Convergence of Artificial Intelligence and
Blockchain in Smart Cities. Smart Cities 2023, 6, 764–795.
17. Oprea, S.-V.; Bâra, A. An Edge-Fog-Cloud computing architecture for IoT and smart metering data. Peer-to-Peer Netw. Appl.
2023, 16, 818–845.
18. Zrelli, I.; Rejeb, A. A bibliometric analysis of IoT applications in logistics and supply chain management. Heliyon 2024, 10,
e36578.
19. Samsul, S.A.; Yahaya, N.; Abuhassna, H. Education big data and learning analytics: A bibliometric analysis. Humanit. Soc. Sci.
Commun. 2023, 10, 709.
20. Lundberg, L. Bibliometric mining of research directions and trends for big data. J. Big Data 2023, 10, 112.
21. Abdian, S.; Khadivar, A. A Bibliometric Analysis of Research on Big Data and Its Potential to Value Creation and Capture. Iran.
J. Manag. Stud. 2021, 16, 1–24.
22. Wang, J.; Kim, H.-S. Visualizing the Landscape of Home IoT Research: A Bibliometric Analysis Using VOSviewer. Sensors 2023,
23, 3086.
23. Wang, Y.; Zhang, F.; Wang, J.; Liu, L.; Wang, B. A Bibliometric Analysis of Edge Computing for Internet of Things. Secur. Com-
mun. Networks 2021, 2021, 5563868.
24. MGe, M.; Bangui, H.; Buhnova, B. Big Data for Internet of Things: A Survey. Futur. Gener. Comput. Syst. 2018, 87, 601–614.
Sensors 2025, 25, 906 37 of 38
25. Alshehri, F.; Muhammad, G. A Comprehensive Survey of the Internet of Things (IoT) and AI-Based Smart Healthcare. IEEE
Access 2020, 9, 3660–3678.
26. Guo, Y.-M.; Huang, Z.-L.; Guo, J.; Li, H.; Guo, X.-R.; Nkeli, M.J. Bibliometric Analysis on Smart Cities Research. Sustainability
2019, 11, 3606.
27. Šajnović, U.; Vošner, H.B.; Završnik, J.; Žlahtič, B.; Kokol, P. Internet of Things and Big Data Analytics in Preventive Healthcare:
A Synthetic Review. Electronics 2024, 13, 3642.
28. Misra, N.N.; Dixit, Y.; Al-Mallahi, A.; Bhullar, M.S.; Upadhyay, R.; Martynenko, A. IoT, Big Data, and Artificial Intelligence in
Agriculture and Food Industry. IEEE Internet Things J. 2020, 9, 6305–6324.
29. Keleko, A.T.; Kamsu-Foguem, B.; Ngouna, R.H.; Tongne, A. Artificial intelligence and real-time predictive maintenance in in-
dustry 4.0: A bibliometric analysis. AI Ethics 2022, 2, 553–577.
30. Dai, Z.; Zhang, Q.; Zhu, X.; Zhao, L. A Comparative Study of Chinese and Foreign Research on the Internet of Things in Edu-
cation: Bibliometric Analysis and Visualization. IEEE Access 2021, 9, 130127–130140.
31. Garcés-Giraldo, L.; Patiño-Vanegas, J.; Espinosa, R.; Benjumea-Arias, M.; Valencia-Arias, A.; Lampen, M.C. Internet of Things—
IoT research trends from a bibliometric analysis. J. Inf. Syst. Eng. Manag. 2023, 8, 2468–4376.
32. Liang, C.; Shah, T. IoT in Agriculture: The Future of Precision Monitoring and Data-Driven Farming. Eig. Rev. Sci. Technol. 2023,
7, 85–104.
33. Fasciolo, B.; Panza, L.; Lombardi, F. Exploring the Integration of Industry 4.0 Technologies in Agriculture: A Comprehensive
Bibliometric Review. Sustainability 2024, 16, 8948.
34. Munirathinam, S. Chapter Six—Industry 4.0: Industrial Internet of Things (IIOT). Adv. Comput. 2020, 117, 129–164.
35. Din, I.U.; Guizani, M.; Hassan, S.; Kim, B.-S.; Khan, M.K.; Atiquzzaman, M.; Ahmed, S.H. The Internet of Things: A Review of
Enabled Technologies and Future Challenges. IEEE Access 2018, 7, 7606–7640.
36. Purwanto, H.; Ahmad, R.; Dirgantari, P. The Role of the Internet of Things (IoT) in Business and Marketing Areas: A Systematic
Literature Review Using the Bibliometric Analysis Approach. In Proceedings of the 5th Global Conference on Business, Man-
agement and Entrepreneurship, West Java, Indonesia, 18 August 2021.
37. Valencia-Arias, A.; Dávila, J.R.; Londoño-Celis, W.; Palacios-Moya, L.; Hernández, J.L.; Agudelo-Ceballos, E.; Uribe-Bedoya, H.
Research Trends in the Use of the Internet of Things in Sustainability Practices: A Systematic Review. Sustainability 2024, 16,
2663.
38. Oprea, S.-V.; Bâra, A.; Puican, F.C.; Radu, I.C. Anomaly Detection with Machine Learning Algorithms and Big Data in Electricity
Consumption. Sustainability 2021, 13, 10963..
39. Dian, J.F.; Vahidnia, R.; Rahmati, A. Wearables and the Internet of Things (IoT), Applications, Opportunities, and Challenges:
A Survey. IEEE Access 2020, 8, 69200–69211.
40. Pachouri, V.; Pandey, S.; Gehlot, A.; Negi, P.; Chhabra, G.; Joshi, K. Agriculture 4.0: Inculcation of Big Data and Internet of
Things in Sustainable Farming. In Proceedings of the IEEE International Conference on Contemporary Computing and Com-
munications (InC4), Bangalore, India, 21–22 April 2023; pp. 1–4.
41. Murugesakumar, B.; Dharshini, D.; Dinesh, K. IoT and Big Data in Agriculture Revolutionizing Farm Management and Produc-
tivity. In Proceedings of the 7th International Conference on Electronics, Communication and Aerospace Technology (ICECA),
Coimbatore, India, 22–24 November 2023.
42. Majid, M.; Habib, S.; Javed, A.; Rizwan, M.; Srivastava, G.; Gadekallu, T.; Lin, J.-W. Applications of Wireless Sensor Networks
and Internet of Things Frameworks in the Industry Revolution 4.0: A Systematic Literature Review. Sensors 2022, 22, 2087.
43. Eirini; Georgios, B.; Maria, L.; Giorgos, V.; Athanasios, A.; Elpiniki, P.; Dionysis, B. Smart Technologies for Sustainable Water
Management: An Urban Analysis. Sustainability 2021, 13, 13940.
44. Yi-Kyung, H.; Ze-Yu, W.; Young, C.J. Global Research Trends on Smart Homes for Older Adults: Bibliometric and Scientometric
Analyses. Int. J. Environ. Res. Public Health 2022, 19, 14821.
45. Jianfeng, L.; Xiao, C.; Hwanyong, K. Mapping a Decade of Smart Homes for the Elderly in Web of Science: A Scientometric
Review in CiteSpace. Buildings 2023, 13, 1581.
46. Tan, H.; Othman, M.H.D.; Kek, H.Y.; Chong, W.T.; Nyakuma, B.B.; Wahab, R.A.; Teck, G.L.H.; Wong, K.Y. Revolutionizing
indoor air quality monitoring through IoT innovations: A comprehensive systematic review and bibliometric analysis. Environ.
Sci. Pollut. Res. 2024, 31, 44463–44488.
47. Zhang, F.; Chan, A.P.; Li, D. Developing smart buildings to reduce indoor risks for safety and health of the elderly: A systematic
and bibliometric analysis. Saf. Sci. 2023, 168, 106310.
48. Magara, T.; Zhou, Y. Internet of Things (IoT) of Smart Homes: Privacy and Security. J. Electr. Comput. Eng. 2024, 2024, 7716956.
Sensors 2025, 25, 906 38 of 38
49. Ahmed, A.; Belrzaeg, M.; Nassar, Y.F. A comprehensive review towards smart homes and cities considering sustainability de-
velopments, concepts, and future trends. World J. Adv. Res. Rev. 2023, 19, 1482–1489.
50. Khan, S.; Ali, H.; Shah, Z. Systematic analysis of smart homes: Current trends and future recommendations. Cogent Eng. 2024,
11, 2344452.
51. Scala, Á .; Aguilar Cuesta, M. Rodríguez-Domenech and M. Cañizares Ruiz, “Bibliometric Study on the Conceptualisation of
Smart City and Education. Smart Cities 2024, 7, 597–614.
52. Rejeb, A.; Rejeb, K.; Abdollahi, A.; Keogh, J.G.; Zailani, S. Smart city research: A bibliometric and main path analysis. J. Data
Inf. Manag. 2022, 4, 4.
53. Gupta, A.; Gupta, S.; Memoria, M.; Kumar, R.; Kumar, S.; Singh, D.; Tyagi, S.; Ansari, N. Artificial Intelligence And Smart Cities:
A Bibliometric Analys. In Proceedings of the 2022 international conference on machine learning, big data, cloud and parallel
computing (COM-IT-CON), Faridabad, India, 26–27 May 2022.
54. Hajoary, D.; Narzary, R.; Basumatary, R. Mapping the Landscape of Smart City Research: A Bibliometric Analysis. In Smart
Cities: Innovations, Challenges and Future Perspectives; Springer: Berlin/Heidelberg, Germany, 2024; pp. 83–112.
55. Puliga, G.; Bono, F.; Gutiérrez Tenreiro, E.; Strozzi, F. Bibliometric Analysis of Scientific Publications and Patents on Smart Cities;
Publications Office of the European Union: Luxembourg, 2023.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual au-
thor(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
