Navigating the Digital Public Sphere: An AI-Driven Analysis of Interaction Dynamics across Societal Domains

Abstract

The increasingly digital nature of modern societies necessitates continually examining how individuals interact in the public sphere. This systematic literature review comprehensively analyzes emerging research on public interaction across diverse contexts. By employing an innovative method of applying artificial intelligence on a large-scale academic corpus, we systematically identified and categorized eight major research clusters: social media and public discourse; public Governance in health and education; urban environments and data systems; group interaction dynamics; complex systems modeling; human-display interfaces; political processes; and public service design. Sub-topic mapping revealed key themes such as digital civic engagement, transport sustainability, behavioral dynamics, and socio-environmental impacts. Our interdisciplinary synthesis highlights public interaction as a multifaceted phenomenon intertwined with technological change, policy decisions, environmental factors, and social constructs. These insights underscore the need for holistic, cross-disciplinary approaches to navigate the challenges and opportunities of public interaction in our rapidly evolving digital age. This review provides a unified knowledge base to guide future research while informing decision-makers on cultivating participatory, adaptive, and sustainable public spheres.

1. Introduction

The dynamic nature of our world, shaped by transformative events such as the COVID-19 pandemic [1], technological advancements exemplified by smart city systems [2], and the escalating interconnectedness prompted by globalization [3], necessitates an ongoing scrutiny of our modes of public interaction. With the emergence of the metaverse, a virtual public sphere, public and interaction studies have gained significant relevance [4]. In this context, integrating public screens for disseminating important public information, advertisements, and entertaining content into our public spaces has become a growing phenomenon.

Exploring public sphere interactions forms a critical component of social research and can be accessed via manual field searches. Nevertheless, a comprehensive overview of ongoing research into public sphere interactions is essential, given the continually evolving influencing factors in contemporary dynamic societies.

Societies are dynamic entities with complex social structures and cultural norms, and understanding how individuals interact within the public sphere is of utmost importance for social scientists, policymakers, and community leaders [5,6,7]. The escalating ubiquity of information has become a cornerstone of the public sphere, significantly shaping social dynamics and communication patterns within societies.

To provide a comprehensive understanding of the diverse field of public interaction research, this work will present an overview of its various subtopics, providing valuable insights for societies to navigate and adapt to the digital age. Considering the prevalent subject of public sphere interaction, which has generated numerous papers that can be overwhelming to read individually, we have chosen a specific approach to address this challenge [8].

The scientific community highly values review articles, as they are essential references for studies that influence societal development. Ongoing discussions revolve around the methodologies for producing high-quality review papers, and by incorporating artificial intelligence in the review process, we can enhance the accessibility and relevance of research findings for societies and their decision-making processes [9,10].

We adopt an innovative approach in this paper to generate a comprehensive overview of research directions in the public sphere and their interaction. Instead of relying solely on manual analysis, we employ a new methodology called the STIRL (systemic taxonomy for information retrieval in the literature) method by Buchkremer et al. for generating reviews [8,10]. By embracing this cutting-edge technique, we advance our understanding of various societal issues and pave the way for future researchers to explore the broad implications of public-sphere interaction in societies. Through this research, we aim to contribute valuable knowledge and understanding to public-sphere and interaction research, empowering societies to make informed decisions and fostering positive social change.

2. State of Research

The growing ubiquity of information has become a cornerstone of the public sphere, significantly shaping social dynamics and communication patterns within societies. Several review articles encompass various aspects of public sphere interactions. We have selected these broader reviews to compare the results with our cluster analysis. This comparative approach aims to juxtapose manually curated reviews with our systematic clustering methodology, thereby offering a comprehensive understanding of the state of research on public interactions.

Pandel and Pant’s article identifies various public living space topics from 61 analyzed articles and focuses in particular on the central city of Nepal, Kathmandu [11] The following main subjects can be mentioned:

• Physical design and infrastructure
• Accessibility and reachability
• Safety aspects
• Social integration
• Political and administrative barriers
• Socio-economic factors

This article provides a detailed examination of the elements that influence the vitality of public spaces and offers practical recommendations for their planning and design. By incorporating these perspectives, our research further delves into the multifaceted nature of public spaces, aiming to augment the understanding and enhancement of public living environments.

Similarly, another article investigating public spaces through a systematic literature review of 64 papers identified the following key topics [12]:

• Design and use of public spaces:
• Accessibility and inclusivity
• Safety and perception
• Community participation
• Political obstacles
• Use of technologies such as geographic information systems and social media
• Need for further studies on the long-term impact and sustainability of public spaces

The authors of this review underscore the necessity of an integrated and interdisciplinary approach to comprehending and enhancing various aspects of the design, use, and management of public spaces. By combining these perspectives, our research delves deeper into the multifaceted nature of public spaces, enhancing understanding and informing strategies for the improvement and sustainable management of public living environments.

Furthermore, the article “Behavior in Public Open Spaces: A Systematic Review of Studies with Quantitative Research Methods” offers an analysis of research that focuses on the behavior of people in public open spaces [13]. This review identified the following relevant research subjects from the 116 studies analyzed:

• Use and activities in public open spaces
• Health
• Safety and perception
• Influence of physical design on behavior
• Social interactions and community building
• Environmental impact and sustainability
• Emphasizing the need for interdisciplinary approaches and the integration of qualitative methods to complement quantitative findings

The authors emphasize the importance of well-designed public open spaces for the social life and well-being of the urban population. They highlight the need for further research to understand better the diverse aspects of the use and design of such spaces.

The three reviews presented reveal both different and similar topics. While they have distinct focuses, they all address the public sphere, similar to our review. The following topics were identified by at least two of the reviews:

• Physical design
• Safety and perception
• Social interactions and communities
• Politics
• Sustainability

In the discussion, we will refer to the articles presented here and compare which topics were also identified by our methodology and what the differences are. It will enable us to recommend the extent to which our approach can offer advantages in application.

3. Materials and Methods

The present study employed the method proposed by Buchkremer et al. in 2019, known as the systemic taxonomy for information retrieval in the literature (STIRL) method [8]. This approach combines literature analysis with the support of artificial intelligence, enabling researchers to evaluate the content of any number of articles. Furthermore, vom Brocke et al.’s literature review process, introduced in 2009, was also incorporated into our research [10].

The STIRL method, as applied in this study, has proven effective for conducting a comprehensive analysis of a wide range of articles and their respective content. Its usage has been previously demonstrated by Weber and Buchkremer in a 2022 study about blockchain-based reference architectures [14]. This study is a practical example showcasing the combined utilization of Buchkremer’s STIRL system and vom Brocke’s literature review methodology.

By integrating these established methodologies, our research aims to advance the understanding of the subject under investigation while harnessing the benefits of artificial intelligence to enhance the efficiency and effectiveness of the literature review process. Using Buchkremer et al.’s STIRL method and vom Brocke et al.’s literature review approach provides a robust foundation for our study, improving our findings’ overall quality and reliability.

Starting with Phase I of vom Brocke’s review cycle, we needed to define the review scope. We apply Cooper’s taxonomy, which classifies six aspects of a planned field of references [15]. The first one is the focus. This review may include research methods, theories, and applications, as we want a comprehensive overview of different topics within one parent topic. The goal is to filter out the central issues of the research field. As we group the literature, it is a conceptual organization in this study. Moreover, we do not pursue an espousal of position but a neutral representation as perspective.

Since this is a review of a subject area that can be limited to the social sciences, our audience is specialized scholars. The last characteristic of Cooper’s taxonomy is the coverage. We restrict our literature to the article type and look closely at particular articles in the following, so we must refer to an exhaustive and selective coverage.

First, a literature corpus must be created, which we built on the online databases shown in

Table 1. Literature databases.

Search String	Database	Hits	Filter
(“Document Title”:public) AND (“Document Title”:interaction)	IEEE	47	Conferences, Journals
Title: public] AND [Title: interaction]	Science Direct	107	Research articles
Title: public] AND [Title: interaction]	Sagepub	111	Research article
Title: public] AND [Title: interaction]	ACM	146	Research article
“public” in title and “interaction” in title	Wiley	149	Journals
	Total	560
	Total after data cleansing	546

Date of data collection: 21 January 2023.

. We tested different academic search engines and chose the ones with the highest number of hits. The search string was created by entering various filter options with the terms “public” and “interaction”. The results best matched the searched topic we got with the filter that looked for “public” and “interaction” in the title of the articles. Moreover, we selected only research articles or journals and conferences depending on the possible filter options in the databases.

It is essential to highlight that we deliberately employed a simple search algorithm that focused solely on the titles of articles. This approach was chosen due to the broad nature of the subject under investigation. We aimed to gather a representative quantity of relevant papers rather than attempting to include every article available. By focusing our search on the titles of articles, we sought to efficiently capture the essence and relevance of various research contributions within the field. This targeted approach allowed us to streamline the literature search process, ensuring that we obtained a representative selection of papers that best addressed the specific aspects of our study. Considering the vastness of the subject, it would have been impractical and unnecessary to include every single article available.

By employing this selective strategy, we aimed to balance inclusivity and relevance, ensuring that only the most pertinent and meaningful papers were considered for our research.

Therefore, our deliberate use of a simple search algorithm and limited scope to the titles of articles were conscious decisions made to optimize the gathering of a representative volume of relevant research papers, eliminating unnecessary duplication and ensuring the focus remains on the most significant contributions to the field.

The downloaded data, saved in “Mendeley1”, contained the following information:

• Type (journal article, conference paper, etc.);
• Publication year;
• Title;
• Author(s);
• Publication title (name of the journal or conference, etc.);
• Abstract.

We curated a corpus of 546 references by removing redundant and irrelevant data, such as duplicates and missing datasets. This rigorous data cleansing process is crucial as it ensures the quality and reliability of our research findings, thereby enhancing the credibility of our study.

Our data underwent a meticulous preprocessing phase, where we word-stemmed it into its original form and implemented a publicly available stop-word list [13]. This approach ensures no erroneous results due to different word variants. For instance, ‘public’ and ‘public’ are treated the same as ‘public’. The stop-word list guides the system to disregard frequently occurring but irrelevant words such as ‘and’. This thorough preprocessing sets the stage for the subsequent analysis and use of artificial intelligence.

The STIRL methodology, a systematic and rigorous approach, was instrumental in fulfilling the goals of phases I, II, and III proposed by vom Brocke 2009. These steps include defining the scope, conceptualizing the appropriate search string, and conducting a comprehensive literature search, as described above.

Phase III also includes a backward and forward search and evaluates references. Vom Brocke does not specify how many studies should be taken for a broad overview of a research field. This part was done manually by presenting three articles that matched the respective Cluster to a high degree and with three Q1 ranking articles according to the Scimago Journal & Country Rank. While the Q1 ranking does not guarantee the highest quality, and there are indeed high-quality papers with lower or no Scimago ranking, we increase the likelihood that the articles in question are highly qualified [16].

4. Results

The abstracts were classified using topic modeling via latent Dirichlet allocation (LDA). The literature classification results are word clusters, each describing one topic in the data corpus, i.e., the article’s contents [10]. Our analysis identified nine clusters, which can be seen in Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8 and Figure 9. The counting of the clusters starts at zero, so we will keep this in mind in the following. Each article in the corpus was assigned to only one Cluster.

Ten words succinctly describe the clusters, as depicted in Figures in the Section 4.1, Section 4.2, Section 4.3, Section 4.4, Section 4.5, Section 4.6, Section 4.7, Section 4.8 and Section 4.9 To generate short titles, we leveraged ChatGPT 4.0 based on the frequency of words in a given cluster. We then entered the terms and frequencies in our prompts and requested a list of titles. The following shows the exact list given by ChatGPT:

Topic 0 (Social Media and Public Interaction)

• “Social Media Dynamics and Public Discourse”
• “Online Public Interactions and Political Communication”
• “Media Influence on Public Participation”

Topic 1 (Urban Spaces and Interaction)

• “Urban Information Systems and Human Interaction”
• “Data-Driven Spaces: Urban Environments and Public Participation”
• “Public Spaces and Technological Integration in Urban Contexts”

Topic 2 (Health, Education, and Governance)

• “Public vs. Private Governance in Health and Education”
• “State Governance and Educational Research: Public Health Implications”
• “Governance Systems in Health and Learning Environments”

Topic 3 (Human Technology Interaction)

• “Interactive Technologies and Public Engagement”
• “Human-Technology Integration in Public Spaces”
• “Technological Approaches to Public Interaction”

Topic 4 (Group Dynamics and Risk Studies)

• “Implications of Group Dynamics in Public Interactions”
• “Risk and Group Behavior in Public Interaction Studies”
• “Social Group Dynamics and Risk Perception”

Topic 5 (Governance and Environmental Studies)

• “Public Governance and Environmental Factors”
• “Local Government and Environmental Impact Analysis”
• “Interaction of Public Governance with Environmental and Social Factors”

Topic 6 (Public Display Interaction)

• “Interactive Public Displays and User Experience”
• “Design and Interaction with Public Digital Displays”
• “Public Display Technology: Interactive Design and Content Management”

Topic 7 (Policy and Public Affairs)

• “Public Policy Interactions and Governmental Transparency”
• “Political Systems, Public Policy, and Social Impact”
• “Public Affairs: Policy Making and Government Transparency”

Topic 8 (Public Service Management)

• “Public Service Systems: Interaction and Management”
• “Service Systems in Public Governance and Interaction”
• “Public Service Process Optimization and Systems Management”

Through meticulous scrutiny and assessment of the suggested titles, particularly about the top three abstracts, we selected the title that best aligned with our estimation of appropriateness or created one on our own if needed. This careful selection process ensures the accuracy and relevance of our research findings. The titles have been used for Topics 6, 7, and 8 as proposed by the system. We used different words from the suggestions for topics 0, 1, 2, and 4 but stuck close to them.

Only the titles for topics 3 and 5 have been strongly modified based on the top three abstracts described in the corresponding subchapters in the following.

We have now obtained our initial findings, providing an overview of issues related to public interaction. Utilizing the recommended forward-backward search method, we analyzed the top three articles from each Cluster to examine the topics covered within their content [10]. We present three reputable articles per Cluster to supplement the topic area and form a topic map. At the end of each cluster description, we conclude with the significance of the results found for society and public interaction research.

4.1. Topic 0: Social Media: Public Interactions and Research

In Figure 1, the ten descriptive keywords of Cluster 0 are visualized.

The high frequency of “social”, “public” and “media” indicates a social media platform focus in this Cluster.

In the following section, we will present the three articles with the highest percentage matching the Cluster and three additional papers not included in the Cluster to provide an overview of this field.

4.1.1. Top Three Cluster Articles

Toepfl and Piwoni’s research departs from previous studies by applying counter-public theory to examine online political discourse [17]. It focuses on comments posted on news websites in response to articles about a new anti-euro party in Germany. The study compares the framing of articles on different news websites with the content of comments posted under these articles. Different counter-publics are identified in the comment sections of right-wing, left-wing, tabloid, and non-tabloid newspapers. The effects on democracy are examined in the analysis.

Cohen’s article examines the relationship between the president’s civil rights agendas and the public [18]. It explores how presidential leadership influences the public’s civil rights priorities and vice versa. The focus is on understanding the dynamics and interactions between the two agendas and illuminating how presidential leadership and representation shape civil rights discourse and policy.

Kaplan and Kupper propose an approach to studying media events, focusing on reality TV shows such as Big Brother [19]. They propose an ‘interaction-centered’ perspective, examining how the interactions between the participants and the media environment contribute to the audience’s overall experience. In the case of Big Brother, they are likely to examine how the show creates a sense of intimacy between the participants and the audience through media interactions and how these influence perceptions of reality and authenticity in media consumption.

4.1.2. Additional Articles

While searching for highly ranked journal articles about Social media: public interactions and research, we found the following three articles that highlight additional aspects of the field.

Kreijns et al. address challenges in collaborative online learning environments and focus on the need to consider social interaction and community solidarity [20]. It investigates factors that impact successful group dynamics. They discuss social interaction pitfalls and suggest recommendations based on educational research and teacher knowledge.

The article Mergel presents a structured framework for understanding and analyzing social media interactions in the public sector [21]. It can offer guidelines or methods for interpreting different interactions, such as comments, shares, or likes, on social media platforms used by government or public sector organizations. The goal is to help public sector practitioners better understand and use social media for public engagement, communication, and feedback.

Zheng and Zheng explore how social media is used as a platform for innovation in the public sector [22]. It could examine how the dissemination of information and interactions facilitated by social media channels contribute to innovation processes within government or public organizations. The focus could be on understanding the mechanisms through which social media platforms are used to generate insights, foster collaboration, and develop innovative ideas in the public sector.

4.1.3. Key Findings

From our analysis, it appears that Cluster 0 includes various sub-themes primarily concerned with the interaction between the public and multiple aspects of the public sector. These topics range from the use of social media in government institutions to online learning environments and the role of social media in promoting innovation in the public sector. The importance of counter-publics in online policy discourse and the interaction between participants in media events and the audience is also examined. Figure 2 visualizes the found topics.

One conclusion from these topics is that digital platforms and technologies increasingly shape the interaction between the public and the public sector. Social media plays a central role in this and is used for both informal interactions and formal purposes such as political discourse and innovation processes. The study of counter-publics indicates that different voices and perspectives in the digital sphere can influence the public discourse climate.

These research topics indicate that digital interaction between the public and the public sector is increasingly important in shaping policy, Governance, and social change. They also show how people interact with each other and how state institutions are evolving in the digital age.

4.2. Topic 1: Data and Information in Urban Environments

Figure 3 shows which ten keywords are most representative of cluster 1.

The most frequent keywords in Cluster 1 focus on how data and information are involved in people’s interactions. Moreover, the Cluster could be centered on urban environments.

The following will present the three most relevant articles in the analysis.

4.2.1. Top Three Cluster Articles

Ayaz et al.’s primary focus is to collect comprehensive data on drug information from public sources, with particular interest in developing such source-based details [23]. The study seeks to locate, gather, and scrutinize all available data sets to understand possible drug-drug interactions fully. The primary goal of this database would be reliability: to be a reliable resource for healthcare professionals and researchers (as well as other stakeholders) to use in their work of assessing drug risks and preventing adverse drug interactions.

Byrd delves into adopting social exchange theory to view client-nurse dealings amidst public maternal and child home health care visits [24]. Qualitative research seeks to comprehend the nature and quality of such client-nurse dynamics. The focus is on how these exchanges impact resource sharing, information flow, and mutual support that foster maternal and child well-being.

The study of Tammentie et al. developed a theory about the relationship between public health nurses and a mother who got depression after the birth of her child [25]. The study aimed to understand and analyze the views of family members, particularly mothers, on their interactions with health professionals. The authors investigated how these interactions affect the well-being of mother and child and what support is available to families to cope with postnatal depression.

4.2.2. Additional Articles

Up to this point, the articles have all been related to medical topics. Three Q1 articles from the Cluster’s subject area are presented below, which shed light on further issues and look at the field from other perspectives.

Heikinheimo et al. approach the use of urban green spaces by looking at the geographical information created by the users [26]. This article can detail people’s attitudes and behaviors regarding using green spaces in urban areas. It can do this through available geographical information such as geolocation data, photos, and comments people make on social media and various platforms. The analysis will use the findings and/or representations to influence the planning and design of urban green spaces for better use and satisfaction.

The article from Miller and Small discusses one of the most effective tools and technological innovations to ensure remote sensing’s potential in urban environmental research and policy [27]. The paper begins with a literature review and a methodology of explaining and analyzing remote sensing’s application to studying urban ecological topics, such as air and water quality, urban vegetation coverage and land use, and urban heat islands. The given paper aims to realize the advantages and disadvantages of using remote sensing for implementing urban-environmental research and policy, as well as to provide recommendations for further research and policymaking.

Firstly, Dembski et al.’s article focuses on urban digital twin usage in smart cities and for citizens [28]. More specifically, the authors observe the whole process within the case study on Herrenberg. The urban digital twin is a computer technology that generates a virtual model of the actual city by collecting relevant data from different sources. The article analyzes Herrenberg’s featured use of urban digital twins as instruments to enhance urban planning, infrastructure, and quality of life management, along with smart citizens’ engagement in these processes. Furthermore, the paper presents an overview of experts’ and citizens’ opinions about the main challenges, opportunities, and likely implications for city services and inhabitants, concluding with future research areas.

4.2.3. Key Findings

Figure 3 summarizes five topics related to cluster data and information in urban environments. First, we presented a study dealing with publicly available information on medicines. We have mapped this field about public health and medical data. Similar studies on other useful medical information that would be helpful to the public are conceivable. The interaction of citizens with public healthcare facilities and their employees also falls under this first topic. In addition to the studies presented here, using technologies and digital platforms could also be interesting.

Urban green spaces are the next topic area we have found. As presented, it deals with citizens’ use of such areas. On the other hand, the influence of such green spaces in cities on health and climate change can also be interesting.

The study presented on green spaces in urban areas proposed using user-generated geographical data. We have added this topic as a further subtopic, as it could be applied to more than just green spaces. Social media, mobile applications, or platforms can be used to gain insights into the behavior and needs of the population.

In remote sensing, researchers are concerned with using technologies and data to monitor environmental factors that can be used for scientific and political purposes.

Finally, we identified the topic of using digital twins. The creation of digital models of urban environments can be used in various ways to anticipate changes and create different scenarios in planning. The interaction of infrastructure and citizen behavior can be analyzed and optimized.

The variety of subjects looked into within public interaction in urban areas demonstrates that a holistic strategy is required to tackle the issues and possibilities related to city life. The findings from such studies could make urban living spaces more efficient, healthy, green—and suited to the needs of the people who inhabit them. In turn, this kind of adjustment could then enhance the quality of life and wellness throughout society—as a whole.

4.3. Topic 2: Public vs. Private Governance; Health and Education

Figure 5 highlights that the research in Cluster 2 could address the Governance of health and education systems.

4.3.1. Top Three Cluster Articles

The paper penned by N. Zamora, E. August, E. Fossee, and O. S. Anderson presents an article on a study that delves into the impact of remote learning for graduate students in public health interactions and feelings of belongingness [29]. It questions how the transition from physical to virtual classrooms influences students’ attachment to the materials, their peers, and even professors—curious about its impact on those who share that same virtual space as them every other day.

Furthermore, it looks at perception from changes such as this—it seeks now to paint a picture surrounding the sense that students have belongingness to an academic community. The paper aims to explore issues surrounding challenges and opportunities unveiled through remote learning for graduate education, especially those undertaking public health programs, seeking more insight into what these programs might offer or lack due to the limitations imposed by physical presence during learning sessions.

Hashash et al. analyzed teacher-student interaction in elementary schools in Lebanon, focusing on Grade six classes [30]. Based on their findings, they provide suggestions for designing classrooms. The frame of reference here is symbolic interactionism. The paper discusses a sociological theory that underscores the importance of symbols and language in constructing social interactions and relationships, how students and teachers exchange information and interpret symbols—what these symbols mean to them, and how they define them based on their engagement.

Tzankova looks at the market-induced control of fisheries and the market-driven monitoring of electricity decarbonization [31]. She concludes that private environmental regulation can positively impact public policies. Her findings suggest that business groups can support public policy as private regulators. This demonstrates how private efforts can enhance more comprehensive environmental regulation: the report underscores the interplay between private Governance and government action. It also underscores how ENGOs oversee ecological concerns—which should be seen as a triangular relationship between private authorities (ENGOs) and government actions regarding oversight of environmental issues.

4.3.2. Additional Articles

After searching for more articles from highly ranked journals on public vs. private Governance, health, and education, we found three articles that further described the research field.

The article by Duerrenberger and Warning details the connection between corruption and education in the developing world, specifically focusing on tertiary education [32]. It looks at how different funding sources—whether public or private—determine the level of corruption within these institutions. Using case studies or data obtained through research, this paper demonstrates corruption across various higher education spheres, including admission processes, faculty recruitment, distribution facilities, and academic integrity.

It also finds out how corruption differs based on where funds come from (public sources, private donors, or tuition fees) and how it has impacted quality and parity for those receiving education in these nations. In general, the article seeks to increase our grasp of the complicated interplay between corruption, sources through which money is obtained, and the field of education that takes place in developing nations, especially within higher education.

Zhang et al. delve into the health conditions and job satisfaction that healthcare staff members experience at the peak of an outbreak [33]. It seeks to understand what leads to a healthy status among healthcare workers and what contributes to their job satisfaction during a critical time like this epidemic.

This research is essential as it has significant implications for health policies and programs that would provide support and protection to those working at the forefront during other such crises in the future. In conclusion, this paper is expected to help significantly future readers in the healthcare workforce and their ability to resist any forthcoming public health emergencies like COVID-19.

The article by Unger, Morales, and De Paepe covers the basic principles of health systems and policy research, majorly centering on appraising the healthcare market [34]. They describe the primary goals that research in this area should aim to achieve—which include understanding the workings of markets for healthcare services and products, their efficiency, effectiveness, and any other identified points where improvement or reform might be needed. Additionally, they elaborate on the methodologies involved when conducting such critical research further. These quantitative methods are like statistics together with econometrics, while interviews or case studies fall under qualitative approaches; it also appears that some fields adopt an interdisciplinary approach drawing from economics or sociology, among others, to provide a more holistic view regarding these markets.

Moreover, they also introduce results from their studies or other literature reviews that highlight the performance of healthcare markets—and the consequences for health system outcomes. It includes examining care accessibility, service quality, cost efficiency, and market equity. In conclusion, the paper suits researchers, policymakers, or even practitioners interested in critically appraising healthcare markets and furthering an evidence-based approach to bettering a health system.

4.3.3. Key Findings

Figure 4 provides an overview of the seven research topics found through our analysis of public vs. private Governance: health and education. Through a forward and backward search, we identified specific themes that further subdivide these areas, such as health policy, public health emergencies, public vs. private funding of higher education, corruption in education in developing countries, private-public governance interactions, and distance education.

Health and education are two primary sectors that directly affect societal well-being and development. Finding specific topics in these areas enables researchers to carry out more specific studies and address the particular challenges and opportunities in these sectors.

In the context of research under “Societies”, this implies a necessity to grasp the dynamics between public and private actors within diverse societal landscapes. Both health and education are quintessential elements in any society; their Governance is intertwined with these actors’ interactions, which substantially impact social growth, equality, and population well-being. By delving into detailed scrutiny and study of these areas, we can gain insights into why societies adopt such organizational formations—hopefully paving the way for their progressive evolution in responsiveness towards citizen needs.

4.4. Topic 3: Public Intersectionality: Transport, Industry, Humans, and the Environment

Cluster 3 is described using words connected to Public intersectionality: transport, industry, humans, and the environment (see Figure 7). The following presents the three most representative articles for that Cluster.

4.4.1. Top Three Cluster Articles

The technical report of EFSA Supporting Publications explores the driving forces that give rise to risks in biological realms associated with animals, plants, and public health [35]. An investigative analysis. They unveil the factors fueling these biological risks and reveal how intricately intertwined animal, plant, and public health risks are. The paper delves into this convoluted interplay, seeking a more precise picture as a result of understanding these complex interactions within this domain.

Alofe et al. describe the determination of endocrine-disrupting potential for industrial chemicals via an integrative approach [36]. Various techniques are used, such as accessing public databases, in vitro exposure, and receptor interaction modeling. The objective is to develop a more holistic view of how industrial chemicals can impact the endocrine system through these effects that could be predicted or assessed.

The paper delves into the setup, traffic communication, and stability of a convoy consisting of SAE Level 2 automated cars on roads that are open to the public [37]. Here, we scrutinize how these vehicles behave and their mutual interactions within the traffic milieu. We aim to unravel more insights into how such systems can be effectively actualized along public highways, seeking out impediments or caveats instability issues.

4.4.2. Additional Articles

The articles presented have illustrated intersectionality in transport, environment, and people. In the following, we present three further articles relevant to the topic area.

The paper “Interactivity-as-product and interactivity-as-process” deals with two concepts of interactivity in media and communication research [38]. One perspective sees interactivity as a product—a property or aspect of media content or platforms. Another views interactivity as a process: it directs attention towards the evolving dynamic interaction processes between users and media. The article explores disparities among these dual notions and their footing in research applications; it also delineates resultant bearings for media usage and production. It, therefore, sheds light on another part of our public sphere and its impact.

Kolpin et al. describe nationwide research conducted from 1999 to 2000 to establish the presence of pharmaceuticals, hormones, and other organic pollutants in the rivers of the United States [39]. The study investigated different types of chemicals found in water samples collected from rivers nationwide, including organic pollutants (commonly introduced into wastewater), which can be sourced through various means.

An effort was made to comprehend the range and whereabouts of these contaminants in U.S. waters and their detrimental effects on the surroundings and the human system—this constituted the primary objective of the investigation.

The paper by Moss et al. narrates the process of crafting scenarios for climate change inquiry and forecasting [40]. It delves into methods through which future scenarios can be enriched to paint a realistic picture of what could transpire in climate-altered developments. The authors of this paper take a bold step in promoting scenario development, which paves the way for a robust stakeholder platform—involving decision-makers, researchers, and even members of the public—through which actors sensitize themselves to challenges from climate impacts. It leads to the generation of appropriate adaptation and mitigation strategies due to proper information drawn from these scenarios, thus informed by such instances.

4.4.3. Key Findings

Figure 8 illustrates the research fields related to Public intersectionality: Transport, industry, humans, and the environment.

The topics listed—climate change, pollution, media communication research, transportation and the automation industry, and chemicals’ impact on humans alongside biological threats to public health—depict the intricacy among various societal actors and systems.

As a result of this work, it is suggested that public engagement research demands a multidisciplinary approach with an integrative analysis: viewing technology holistically along with industry needs and environmental impact on society. These insights transcend any field but inform what studies about societies should encompass. Therefore, the main point of this study calls for more holistic scientific investigations without being limited to only one field; they are far broader.

In social research, this means capturing the interconnected structure of various social actors, systems, and environmental components. The findings from this area can highlight the challenges and opportunities of this intertwined complexity. Its goal should be to ensure prosperity and sustainability for a society. Uncovering such entanglements requires an interdisciplinary Synthesis: looking at issues through a tapestry that integrates all the interconnected pieces—a synthesis that reveals the contours that define us as members of society.

4.5. Topic 4: Interaction Studies: Implications of Group Dynamics

Cluster 4 contains articles concerning Interaction studies: Implications of group dynamics. The three most matching studies will be analyzed in the first part of this subchapter. Figure 9 shows the Cluster with its ten keywords.

4.5.1. Top Three Cluster Articles

The research, “Free-ranging male northern map turtles use public information when interacting with potential mates”, examines the conduct of male northern map turtles as they engage possible mates [41]. Its main objective is to establish whether these turtles depend on public information when choosing partners.

This research has valuable findings that help us comprehend the mechanics of intra-group animal interactions, where dynamic information flow affects decisional actions and behavioral patterns. Such insight holds implications for unraveling evolutionary paths leading to group behavior, animal reproduction strategies, and possible applications even reaching wildlife population control.

The paper studies how communication and social interaction-induced anxiety impact interleukin-1 beta and cortisol releases when speaking publicly, which is highly stressful [42]. They aim to discover how fear of public speaking and such social interactions impinges on the biological stress response orchestrated by these two secreted substances within the body. Examining the reactivity of these substances during public speaking provides insights for researchers into realizing how anxiety and stress operate biologically and what factors can mold these responses.

Fear of public speaking or social interaction is individual but can influence the interactions among group members and, at times, even determine the general dynamics of the group. High communication apprehensive individuals—because they cannot interact well with other members due to their issues—can create tension or cause misunderstandings within a group.

Lawson et al. are delving into the public health implications of human-bat interactions [43]. They analyze the risks that could lead to human health complications through bat bites and scratches. The paper may fall under “Interaction Studies: Implications of Group Dynamics” if we view interactions between humans and bats as a group dynamic system. While bats are not human groups, their collective behaviors and interactions with humans can be seen as part of a group dynamic structure.

4.5.2. Additional Articles

The three papers covered animal interactions, human-animal interactions, and group communication. Next, we will show three additional articles on interaction studies and group dynamics.

The article “When to choose dynamic vs. static social network analysis” discusses the issue of determining when to employ dynamic versus static techniques of social network analysis [44]. Social network analysis is an approach that examines the structure and dynamics of relationships and interactions between people or groups. Dynamic network analysis may help to understand the development of relationships and interactions within groups over time. In contrast, static analysis may not capture the full complexity of group dynamics.

Researchers Dikker et al. investigate brain-to-brain synchrony to track fundamental dynamic group interactions in the classroom [45]. They examine how students’ brain activity synchronizes during instruction and how this synchrony relates to in-group interactions.

Studying the synchrony between brains during group dynamics will provide a new perspective on how individual brains work in group environments. It can help us understand the complicated network between personal behavior and collective phenomena, which could lead to more effective educational methods.

The following paper contributes to deepening our knowledge about the communication process in tutorial groups, shedding light on the interactions and dynamics that surface through collective learning, which could be very helpful in many ways [46]. The implications are substantial: they concern both the structuring of learning space and its orchestration and foster effective communication and cooperation within educational situations.

4.5.3. Key Findings

The articles found and presented in Interaction Studies: Implications of Group Dynamics highlighted six research topics. Figure 10 gives an overview of the fields.

The examination has discovered various research fields investigating group relationships and dynamics. It implies that various research methodologies are needed to grasp the dynamics of group interactions in diverse contexts. These areas also indicate the multifaceted nature of society and the various means through which different individuals, animals, or groups coexist and impact each other.

4.6. Topic 5: Interactions and Influencing Factors in Complex Systems

First, we present the three articles that fit to the Cluster: Environmental Factors and Government Interaction. Figure 11 shows the keywords for Cluster 5.

4.6.1. Top Three Cluster Articles

The paper of XU, Wu, and Chen delves into guiding public opinion through controlling influence or interactions within the public or social media platforms [47]. It discusses different constraints that can be placed on interactions—like limiting the spread of misinformation, moderating comments, or even making some posts invisible. It looks at how these constraints affect how people form and share their opinions. The article suggests ways to guide public opinion based on these limitations.

Dobay et al. investigate cooperative behavior in digital microbes and what determines it [48]. They consider cell diffusion, cell density, and public goods features. In a digitized world, the researchers use simulations to model these interactions and observe their impact. The outcomes highlight the specific circumstances under which cooperation is favored (or disfavored) among microbes. This study thus seeks to provide insight into the mechanics of cooperative behavior, mainly influenced by environmental factors.

The paper entitled “Digital piracy and firms’ strategic interactions: The effects of public copy protection and DRM similarity” delves into the impact of digital piracy on shaping firm strategies [49]. In particular, it delves into the appropriateness of using public copy protection and Digital Rights Management (DRM) systems with an eye for similarity. These two factors set the tone for how different companies operating within the same digital market share their products and influence competition as they strive to choke piracy. At the same time, they are still competitive—thus, in this light, some new managerial strategies can be adopted in dealing with these issues.

4.6.2. Additional Articles

The three articles mentioned have a common theme: they deal with complex systems’ interactions and strategies. They analyze how various factors and activities influence behavior within these systems. We will present three additional articles describing the Cluster’s topic area.

The study entitled “Urbanization-induced land degradation and socio-economic challenges in peri-urban areas: Insights from Southern Europe” looks at the impact of urbanization on land and communities in the outskirts of cities in Southern Europe [50]. They aim to highlight the environmental effects likely associated with the expansion of urban areas. They also look at socio-economic challenges because people live in such peri-urban regions surrounding emerging urban cores. The growth that hinders efforts to develop local infrastructures by eclipsing a problem viewed from two different angles (development and sustainability) shows us how important it is and leads us to conclude that any remedy should consider the environmental and social dimension alongside urbanization policies.

The paper of Shannon, Markiel, and Ozier unveils Cytoscape, a tool crafted for delving into the visual and analytic realms of biomolecular interaction networks [51]. It articulates its intricacies by delving into how this program amalgamates various biological data to craft all-encompassing models of these networks. Features and functions find their voice in the narrative with an accent on their skill at wrangling intricate datasets. Cytoscape’s real-world applications in biological research pop up—painted as a means to comprehend molecular signaling pathways and unveil potential druggable targets.

The paper “Consensus and cooperation in networked multi-agent systems” examines how agents can concur and act together in a network [52]. It delves into the algorithms and protocols that foster agreement among agents, including evaluating factors regarding cooperation. Factors considered include the structure of networks, as well as communication patterns that influence coordination. Several challenges from coordination are noted; strategies must be adopted to surmount such obstacles to design successful systems. However, the consensus is essential: this paper shows its significance and where it should find applications—robotics, distributed computing, and social networks.

4.6.3. Key Findings

Figure 12 summarizes all the presented articles within a topic taxonomy of the research subfield of Cluster 5.

Studies reveal that natural and artificial systems display intricate, often nonlinear, dynamics. These dynamics are affected by several factors, such as communication, cooperation, competition, and regulation. In any interaction, information flow and resource accessibility, coupled with regulation and structural conditions, are pivotal determinants of the outcomes that transpire from it.

These fields offer important lessons when viewed through the societal lens. Human societies are intricate tapestries of individuals and groups where many social, economic, and technological factors influence relationships and actions. Public opinion manipulation strategies reveal how information and media can sway perception and behavior at a mass level—providing insights on par with studies on cooperation in digital microbes and models for human behavior based on socio-economic contexts. Similarly, understanding digital piracy systems and DRM as dual tools—one for regulation, another for disruption—paints technology in its societal role. Urbanization studies highlight challenges (as well as opportunities) associated with rapid growth leading to transformation within communities’ structure.

4.7. Topic 6: Interaction with Public Displays

In Figure 13, Cluster 6 seems to cover research on interactive public displays and users engagement.

The three articles with the highest match score to cluster six will be described in the following.

4.7.1. Top Three Cluster Articles

Khamis et al. describe a system called EyeScout. This system allows users to interact with large public displays using their gaze [53]. EyeScout uses active eye-tracking technology, enabling interaction regardless of the user’s position or movement. The system aims to provide a seamless and intuitive way of interacting with digital content on public screens. The study describes the technical aspects of EyeScout and evaluates its performance. The results show that EyeScout effectively supports gaze-based interaction in various scenarios and improves the user experience with public screens.

The second paper presents GazeDrone, an eye-based public space interaction system [54]. The user is required only to use their eyes; there are no other systems like this that do not require any special gear. GazeDrone uses a drone embedded with eye-tracking capabilities to trail the direction of the user’s vision—such a setup ensures that communication between the user and digital content in public spaces is intuitive and hands-free. The technical components of GazeDrone are discussed, and an evaluation is done on its performance, which indicates that GazeDrone effectively achieves eye-based interaction from different environments publicly depicted.

The third article from the Cluster introduces TextPursuits, a system that facilitates interaction with public displays through text-based gaze tracking [55]. The mechanism includes a dynamic text presentation calibrating and monitoring the user’s gaze. This innovative approach streamlines the interaction process while increasing accuracy—users can participate in the content displayed on public displays by following the text with their eyes. The paper describes the design and implementation details and the performance evaluation of TextPursuits.

4.7.2. Additional Articles

All three papers focus on innovative approaches in user-public display interaction using eye-tracking. They propose easy-to-use solutions where the user does not need to carry any device or equipment. They are all working with eye-tracking technology for an intuitive, hands-free operation. EyeScout and GazeDrone can be used without wearable devices, and TextPursuits uses moving text for calibration and interaction.

Despite their differing approaches, all approaches try to enhance user experience through uncomplicated interactions.

We already can see that this Cluster contains articles about the interaction with public screens. In the following, we will present three strongly perceived papers from this field to determine which topics seem particularly relevant.

Müller et al. study the extent to which attention to public screens depends on the expectation of the content shown [56]. The researchers found that engagement increases when the recipient expects exciting content. Through a survey, they discovered that eye-catching, colorful, and attractive content will likely attract more attention. However, the findings were not subjected to a practical test.

The control of content on public screens was tested by an optimized algorithm of Bushman and Labrinidis [57]. The set-up consists of a screen at a bus stop that shows content that has to be displayed at a particular time (arrival times) and content where this condition does not exist (Twitter channel). In addition, the algorithm can prioritize content. As research gaps, the authors state that the need for such content control and broad usage must be investigated.

Peltonen et al. examine interaction with CityWall, a large public multi-touch display in Helsinki [58]. The experiment explores social aspects such as crowding, teamwork, games, and conflict management. The study discusses how the display affects public space and offers insights for designing similar systems in urban environments. The researchers point to the effect of social learning that can be generated using a large, shared screen. However, they still see gaps in research on the design of applications for groups on this type of screen.

4.7.3. Key Findings

We identified four subtopics inside the field of Interaction with Public Displays. The topics are highlighted in Figure 14.

The presented approaches try to enhance user experience through uncomplicated interactions. Eye-tracking enables hands-free and natural interactions that focus users’ attention and increase engagement. It can make controlling and navigating content easier and create new social interaction opportunities around public displays. Such innovations can democratize access to information and encourage participation in public discourse.

For societies, these technologies improve information dissemination and communal experience. Public displays controlled by eye-tracking and similar technologies can be used in urban environments, for example, to distribute important information efficiently and accessibly. It strengthens the sense of community and encourages citizens’ active participation in public affairs. In addition, such interactive displays can stimulate social interactions and enhance community connections.

One aspect that might be interesting is to investigate what motivates people to interact with public screens. Basic human needs can play a role here. Artificial intelligence in applications on public screens also offers a broad field of research.

4.8. Topic 7: Political and Public Affairs

The articles in cluster 7 mainly handle political and public topics, as can be derived from the keywords described in Figure 15.

4.8.1. Top Three Cluster Articles

The first paper that matched most was from Dhital et al. [59]. It analyzes a frictional economy’s fiscal and monetary policy relationship with fiat money, bank deposits, and government bonds. The study finds that active monetary policy creates a precise balance and increases consumption inequality. At the same time, active fiscal policy can lead to fundamental imbalances and reduce consumption inequality if the government issues few long- or short-term bonds.

Baldi’s article describes the CoDECS project [60]. An open set of cosmological N-body simulations, the CoDECS project is made accessible to the public for models of interacting dark energy. This projective scenario aims to depict a better visual understanding of the behavior and interactions of dark energy in the universe. The article delves into the methodology and technicalities surrounding the simulations and extends an invitation to all scientists: come and use what we have unveiled.

Although the article is primarily scientific, the public availability of the simulations could be relevant for policymaking in cosmological research. Making data available to the general research community supports transparent and collaborative science, which can influence policy decisions in the science sector.

The third paper from the Cluster explores the relationship between monetary and fiscal policies with central bank transparency and public investment [61]. It points out how central bank transparency weakens or strengthens the effectiveness of both financial and fiscal policies while also revealing that market confidence can be enhanced through apparent efforts by central banks, thus calling for increased coordination. The need for open and transparent communication by central banks and the importance of making public investments is also demonstrated here.

4.8.2. Additional Articles

The following three articles will extend the literature analysis in Policy interactions to highlight topics from recommended journals and have a broader impression of its content.

Schneider and Ingram have written an article of high scientific interest on political influence [62]. This paper explores the complexities of social constructs regarding specific populations, how they impact policy, and the details of decision-making processes within political realms. It reveals how perceptions of different groups drive political action: those viewed positively receive support and resources, while those with negative portrayals face punitive policies. The article uncovers these constructions that, if left unaddressed, would empower differentials and societal disparities.

Bond et al. analyze the dynamics of political influence, specifically the impact of political motivational messages sent to many Facebook users during the 2010 US congressional elections [46]. The study shows that these messages directly impacted users’ political self-representation, information-seeking behavior, and real-world voting actions. Moreover, the influence extended to the messages’ recipients and contacts. It highlights the role of close personal connections, including face-to-face interactions, in the diffusion of political behaviors in online and offline social networks.

The article “When Effect Becomes Cause: Policy Feedback and Political Change” explores political influence, explicitly examining the link between public policies and changes in the policy landscape [47]. It assumes that politics exerts its effect through two primary channels: creating incentives and resources for political players and shaping their perceptions of the political environment through information and keywords. These mechanisms impact government elites, interest groups, and the public.

4.8.3. Key Findings

The identified themes illustrate that public interaction and political decision-making are complex and multifactorial systems. They show how different elements, such as economic policy, scientific knowledge, institutional transparency, social perceptions, and communicative strategies, interconnect. Figure 16 lists all subtopics found in the analysis.

A comprehensive understanding of these dynamic interactions is necessary for societies to develop just and efficient political systems. The interactions between economic policies and social constructs directly influence different population groups’ quality of life and opportunities. Central bank transparency and public investment are critical to citizens’ trust in government institutions. Political motivational messages and policy feedback shape the political climate and citizen participation.

4.9. Topic 8: Public Service Management

The last cluster concerns Public Service Management-related keywords; see Figure 17. The three papers that fit the most are presented in this chapter first, and then high-ranked articles from this research field extend the analysis.

4.9.1. Top Three Cluster Articles

All three articles are thematically related to public transportation. The first looks at the growth of intelligent public transport systems and how they interact with the energy system [63]. The study assesses how modern technologies in the field of public transport can improve their efficiency and sustainability; it also looks at electric vehicles and their place in this grid. It states that a successful outcome calls for coordinated planning between these two systems, thus underscoring the need for intelligent control coupled with real-time data analysis: both are instrumental in ensuring a seamless and practical relationship between these two sectors. Though demanding, the survey points out that such an integration could result in a considerable positive impact on energy efficiency as well as reduction of emissions.

The second publication examines green technologies in public transport and their role in promoting operational efficiency [64]. More specifically, it examines the introduction of electric vehicles and their interactions with energy supply dynamics. The paper outlines the need for synergies in planning transportation and electricity landscapes. It highlights the central role of intelligent control in ensuring that their interaction is seamless and productive. The research findings show that such collaboration can improve energy performance at the system level, leading to lower emissions.

The third paper models service interactions in public transport systems [65]. It aims to enhance public transport management and optimize the different services involved in public transport. It discusses how these measures can improve the efficiency and reliability of the entire system; some techniques for such a model are detailed and illustrated with examples showing their impact on planning and control. Results stress that there needs to be unity between all these factors since each plays its role in ensuring the successful management of a public transport system.

4.9.2. Additional Articles

In the following, we highlight three more articles from Public Service Management published in highly recommended journals.

The article “Continuing the development of the public service logic: a study of value co-destruction in public services” discusses value diminishment within public services [65]. It delves into the relationships of different stakeholders where interactions could lead to a reduction in value. It means that the Public Service Logic (PSL) concept has primarily focused on creating value and is being evolved further due to this analysis. The paper outlines four dimensions that signify causes of value reduction in public services and underlines the importance of widening the PSL concept by considering impairments. Also, the paper takes an ecosystem perspective that involves resource integration from several related actors.

The researchers Trischler and Westman Trischler are exploring the design of public services in a world that is already digitalized [66]. The article stresses user experience and suggests that it ought to be prioritized when developing service designs. It argues for a user-centered approach, which can enhance public service efficiency and satisfaction levels. The article discusses several strategies and techniques that should ensure that digital technologies are purposefully blended into the design of public services.

The paper by Kinder et al. examines the ethical use of AI in local public services [67]. Using examples from Finnish cities, the authors discuss the use of AI and observe which ethical standards are implemented. They conclude that AI’s negative ethical implications can be circumvented as long as citizens play an active role in innovation processes; a new learning framework based on social learning is proposed to improve the ethical evaluation of AI-based public services.

4.9.3. Key Findings

The analysis of the last Cluster and the Public Service Management research field identified six topics, as illustrated in Figure 18. These sub-themes make it apparent that modern technologies and new ideas are at the heart of the design and management of public services. The use of intelligent systems and green technologies increases the effectiveness and sustainability of these services; an ideal approach to managing transport services alongside the public service logic towards a user-centered, all-inclusive front is highlighted.

Technology influences the relationship between the public and public services. It should be considered a progressive action. Besides efficiency, such developments would also foster transparency and citizen involvement. The authorities must win trust in their actions, such as prioritizing user-oriented design and ethical aspects.

4.10. Document Distribution across Clusters

In the following, we describe the distribution of articles across the different clusters. Figure 19 illustrates the document count per Cluster.

Cluster 6 contains the highest amount of articles, followed by Cluster 0. Both Clusters are considerably more significant than the others, which indicates that these included research topics (Interaction with Public Displays/Social media: Interactions and Research) are more prevalent in the academic discourse.

The research fields found in Clusters 7 and 3 might be niche topics that affect a few specialists. Alternatively, another explanation could be for some of the topics they are developing and starting to increase over the next few years. For example, there is currently a solid social focus in Western countries on political developments caused by economic problems, immigration, and wars.

5. Discussion

The article aimed to provide a comprehensive understanding of research on public interaction. We used an innovative methodology called “Systemic Taxonomy for Information Retrieval in the Literature” (STIRL), combined with Artificial Intelligence, to analyze a large corpus of academic articles. We focused on different subtopics within public interaction and their influence on social dynamics.

In Section 2, we presented three review articles created manually and ranked between 61 and 116 articles. We identified the following subjects that the reviews have in common:

• Physical design
• Safety and perception
• Social interactions and communities
• Politics
• Sustainability

Compared to our cluster analysis, politics has been mentioned in several clusters with different focuses, mainly in the political and public affairs clusters. The topic of sustainability can be matched with the subtopics we found: environmental pollution, climate change, and corruption and environmental pollution in the developing world. Social interactions and communities also connect to our findings through the Cluster of social media, public interactions and research, and the subtopics of teacher-student interactions and collective learning. The field of safety and perception is addressed by the subtopics of biological risks to animal, plant, and public health, the effects of industrial chemicals on humans, and climate change because they all affect human feelings concerning the perception of a safe life.

We considered several factors to assess our cluster analysis method and suggested using it instead of a manual literature analysis. An enormous number of articles can be analyzed methodically and effectively with cluster analysis, which enables it to spot patterns and themes that a manual analysis might miss. It also offers a quantitative foundation for recognizing themes and trends and lessens the possibility of bias resulting from subjective evaluations. The approach provides a thorough synopsis of the field of study and facilitates the identification of subtopics and specialized areas needing more investigation.

Nevertheless, manual analysis, which may overlook subtler nuances and qualitative aspects of the study, achieves greater depth and detail than cluster analysis. Unclear or poorly written abstracts could compromise the analysis because the method mainly depends on the quality and information content of the abstracts. To make meaningful conclusions from the interpretation of the data, one must also have a thorough understanding of the research field.

Cluster analysis proves helpful when a thorough overview and in-depth article analysis are needed. It can be used for both preliminary research to identify key themes that can be further investigated in a more comprehensive manual analysis and exploratory studies, which look for broad trends and patterns in the research field.

A manual literature review is preferred when a thorough analysis and literature comprehension are necessary. It is particularly true for small—to medium-sized data sets, for which a manual analysis can be completed within a reasonable amount of time and when examining qualitative aspects, theoretical frameworks, or highly complex methodological details.

If an author wants to capture both aspects, a broad and in-depth overview, combining the presented methods is the most efficient. The start should then be a Cluster analysis followed by a manual analysis to go deeper into details and qualitative aspects of the research fields.

6. Conclusions and Limitations

The study is divided into cluster parts and deals with public interaction research topics.

These clusters involve social media interactions, political discourse, collaborative online learning environments, and the use of social media in the public sector for innovation and stakeholder involvement. The work underscores the significance of people’s behavior within the public sphere, particularly when it becomes more critical due to rising digitalization.

The study findings offer considerable knowledge that can benefit social scientists, policymakers, and local leaders as they find their way around—and adapt to—the digital age. The incorporation of artificial intelligence into the literature review is depicted in the paper as an innovative method that increases the productivity and applicability of research outcomes to human societies, where such results are used for decision-making processes.

The COVID-19 epidemic has led to a reevaluation of public spaces. Considerations include safety, flexibility, and the need for digital solutions that allow people to stay connected even during emergencies while ensuring public services are still provided.

Digitalization has changed public interaction: It introduces new communication methods, improves data collection and analysis, and generates advanced urban planning and management ideas in public service. Social media has emerged as a primary platform for public engagement—influencing political conversations, innovation flows, and communication channels within the public sector.

Themes related to transport, industry, and the environment demonstrate the interconnectedness of various societal factors and the need for holistic approaches in public interaction studies.

The present study also shows limitations in the coverage of the individual subject areas. On the one hand, there is a risk of generalizing the study’s results and overlooking important cultural and contextual influences. The data provide a broad overview of research fields but does not emphasize a cross-cultural comparison or local adaption. Sociocultural factors will have partly shaped the analyzed and presented articles, so the results cannot be applied to other studies without further content.

For example, some countries will have political structures that do not allow or support a specific research topic, like those with a different political system. That would lead to a completely different article occurrence in this field.

Parameters such as the history of a country and its local policies are crucial to reflect how the analysis findings might shift or need adjustments when applied to different settings.

The need for more holistic research should be underscored with a call to future studies that will draw from this work—seeking other unexplored areas related to the implications, among other things, of public interaction in dynamic societies.