Privacy and security in mobile technology: A bibliometric analysis for marketing strategies

“The significance of research in the realms of privacy and security cannot be overstated, as these areas necessitate continuous advancements and innovative solutions” (Lakshmi & Kumar, 2017). In the field of Software Engineering, Ebrahimi et al. (2021) conducted a study on mobile application privacy between 2008 and 2018. Their findings revealed that most current research emphasizes the creation of tools for detecting privacy leaks, while there remains a notable scarcity of studies addressing key aspects such as “privacy requirements,” “privacy policies,” and “user perspectives.”

Dhar and Bose (2020) examined the security challenges inherent in the deployment of IoT devices, including mobile IoT, and proposed an integrated security framework that leverages Zero Trust principles and blockchain technology to enhance identity verification and access control within IoT ecosystems. Various cybersecurity approaches have also been developed to safeguard autonomous vehicles from cyberattacks. Poddar et al. (2024) explored the application of machine learning (ML) and deep learning methods to bolster defense mechanisms in autonomous vehicles, which rely on easily manipulated technologies such as cameras, ultrasonic sensors, and infrared sensors. Similarly, Wang et al. (2021) introduced a model-based framework using multi-sensor data to detect and isolate sensor-based attacks within transportation systems.

Zhang et al. (2022) addressed data privacy and security concerns in agricultural IoT by incorporating cryptographic and blockchain solutions to ensure secure management of agricultural data. Mao et al. (2023) investigated 6G edge network technologies, emphasizing their potential for advanced service delivery while identifying critical unresolved security issues.

A comprehensive reading of the literature suggests that privacy and security concerns are multifaceted and continuously evolving across various technological ecosystems, including mobile applications, IoT devices, autonomous vehicles, and next-generation networks. Despite notable technological advancements, much of the existing work approaches these issues from a predominantly technical perspective, often overlooking critical dimensions such as user expectations, regulatory frameworks, and the strategic implications for marketing. Foundational elements like privacy policies and the alignment of security measures with trust-building strategies remain underexplored. Additionally, while innovations span diverse sectors, limited cross-sectoral synthesis hampers a holistic understanding of how such technologies can be deployed securely and responsibly in consumer-facing contexts. By integrating insights from these disparate domains and linking them to marketing-oriented concerns, this study puts forward an interdisciplinary framework that identifies concrete directions for future research and meaningful guidance for practitioners.

Marketers face significant challenges in adopting technological innovations while addressing key concerns such as consumer privacy, data security, personalization, loyalty, confidentiality, and the detection of data breaches.

Varnali and Toker (2010) emphasized the importance of ensuring mobile user privacy and security. Within consumer policy, mobile marketing raises critical issues due to the unique capabilities of mobile communication technologies, which allow for the identification of individual users. This characteristic introduces considerable risks to the privacy and security of personal information – risks that are further intensified by the increased adoption of mobile marketing among younger consumers (Purwanto et al., 2020).

Sinha et al. (2019) investigated privacy concerns and provided insights into enhancing both security mechanisms and user experience. Similarly, Ullah et al. (2023) analyzed privacy risks associated with targeted mobile advertising and proposed solutions to mitigate these concerns.

This body of research highlights the need for future work focused on the development of user-centered privacy protection tools that promote transparency in data usage and empower individuals to safeguard their personal information in mobile advertising contexts.

Rafieian and Yoganarasimhan (2021) investigated the relationship between mobile in-app advertising, regarded as the leading form of digital marketing, and targeting efficiency associated with privacy concerns and economic results. Considering social media marketing as a segmentation tool, Serrano-Malebran et al. (2023) explore relationship between consumer classifications and variables such as privacy concerns, trust and age. In the field of mobile social networks, Lakshmi and Kumar (2017) consider the privacy protection and security as critical issues requiring constant updates and solution development. The convergence of these studies reveals a growing strategic imperative for marketers to integrate privacy and security considerations into the design of MT and personalized marketing tools. The literature underscores not only the operational risks associated with data misuse and breaches, but also the reputational and trust-related consequences that directly influence consumer engagement and brand loyalty. While multiple technological and behavioral insights are offered, there remains a gap in cohesive frameworks that align privacy-by-design principles with STP strategies. Furthermore, few studies provide actionable models that reconcile personalization with ethical data practices across platforms and demographics. This fragmented landscape suggests a need for more integrative approaches to ensure that MT contribute meaningfully and responsibly to long-term strategic objectives.

In this regard, the study aims to synthesize core insights from the existing body of literature and to highlight the role of MT in shaping marketing strategies, particularly through the lens of STP.

The process of selecting documents from the WoS database is presented in Table A1.

This search query aims to identify articles in English that discuss the intersection of marketing strategies with MT, privacy and security concerns. It combines three separate searches: the first focuses on marketing strategies (including STP strategies), the second explores MT (including ML, IoT, blockchain, recommender systems, and tracking technologies), and the third examines privacy and security. The final search combines these three, filtering for articles only, ensuring the results focus on scholarly research.

In this comprehensive bibliometric analysis, the study strives to map the literature’s temporal evolution and delineate its developmental phases. Furthermore, it identifies and evaluates the most influential journals, pinpointing those that have significantly shaped the discourse in this domain. This research will highlight the most impactful authors, recognizing their contributions to advancing knowledge in the field. Additionally, this analysis uncovers the most cited documents and references, thereby illuminating the seminal works that have laid the foundation for current research and that continue to drive scholarly discussions.

This study employed a bibliometric analysis approach supported by multiple tools. The primary analysis was conducted using Bibliometrix, an R package integrated into R version 4.4.0 and executed within the RStudio environment (Aria & Cuccurullo, 2017). To facilitate accessibility and enhance the visualization of results, the study utilized Biblioshiny, a web-based interface that enables interactive, user-friendly bibliometric analysis and graphical output. For concept maps creation, the study made use of PlantUML (https://plantuml.com/en/), an online diagramming tool. PlantUML allows for the generation of thematic concept maps from simple text-based inputs, offering a clear and visually appealing representation of the study’s conceptual framework.

Table A2 provides a snapshot of the research landscape surrounding MT, privacy, and security, highlighting the field’s rapid growth and international collaboration.

The dataset spans 28 years, from 1996 to 2023, and includes over 1,000 documents from 502 sources. The 21.79% annual growth rate indicates a significant increase in research activity. The average of 18.22 citations per document echoes high impact and relevance of the research within the field. The high percentage of international co-authorships (33.77%) evinces a global collaboration in tackling these complex issues. This suggests an expanding field with considerable international interest.

Figure 1 depicts the annual evolution of publications on MT, privacy, and security for marketing strategies from 1996 to 2023, revealing a noteworthy trend over time. Initially, from 1996 to 2008, the number of articles published each year was minimal, indicating limited academic interest in this intersection. However, a gradual increase starting in 2009 marks the onset of a growing recognition of privacy issues in MT. This period of steady growth, which continued until 2018, reflects a more sustained academic focus. A dramatic surge in publications is evident from 2019 onwards, with the number of articles exceeding 200 annually in 2021 and 2022, showcasing an intensified focus, likely driven by advancements in MT, elevated device usage, and rising privacy concerns. Although 2023 reveals a slight decrease, the publication numbers remain high, suggesting a stabilization or a possible shift in research focus.

Figure 1

Temporal publication evolution.

Scientific production in the field of MT, in relation to privacy, security, and marketing strategies, has grown remarkably since 2017, stimulated by the emergence and convergence of innovative technological concepts such as wireless sensor networks (WSN), IoT, ML, deep learning, and blockchain technology.

Table 1 presents key indicators of journal influence and impact within the research area of MT, privacy, and security. These indicators, including h-index (Hirsch, 2005), g-index, m-index, total citations (TC), number of publications (NP), and the starting year of publication (PY_start), allow researchers to assess the relative influence of these journals within the field.

The journals IEEE Access, IEEE Internet of Things Journal, and Sensors are currently leading the field in terms of impact and research output. Their high h-index, g-index, and total citations demonstrate their significant contributions to the field. Interestingly, the table also highlights the emergence of newer journals like IEEE Communications Surveys & Tutorials and Future Generation Computer Systems, which are quickly gaining traction. This implies a dynamic research landscape where established journals are complemented by rising publications shaping a diverse and evolving research environment.

IEEE Access journal explores the crucial intersection of MT, privacy, and security in the context of IoT, 5G, and Industry 5.0. It covers advanced topics like blockchain, federated learning, and ML, focusing on their applications in enhancing mobile privacy and security. By investigating practical applications in areas like smart healthcare and social distancing, the journal demonstrates the relevance of these technologies to marketing and consumer behavior.

Table 2 presents a detailed analysis of the scholarly performance metrics for three authors: Kim-Kwang Raymond Choo, Enrico Cambiaso, and Ivan Vaccari. Each author’s performance is quantified using several indices: the h-index, g-index, m-index, total citations (TC), number of publications (NP), and the year they started publishing (PY_start).

Kim-Kwang Raymond Choo is a leading researcher in the field of cybersecurity, particularly focusing on the privacy and security issues posed by emerging technologies like IoT and autonomous vehicles. His research explores both theoretical and practical aspects of these challenges. He has developed solutions for cross-domain authentication in drone networks using blockchain, proposed security methodologies for self-driving vehicles, and conducted large-scale analyses of IoT vulnerabilities.

Enrico Cambiaso is a researcher specializing in IoT security, focusing on vulnerabilities in protocols like message queuing telemetry transport (MQTT), and developing novel attack methods. He has demonstrated the effectiveness of a low-rate Denial of Service (DoS) attack, SlowITe, which exploits a weakness in MQTT to disrupt service. His research highlights the importance of understanding and mitigating security risks in IoT environments, particularly for critical infrastructure and sensitive data transmission.

Ivan Vaccari is focusing on security vulnerabilities in IoT networks, precisely targeting the MQTT protocol. He has identified weaknesses in MQTT that allow clients to manipulate server behavior and has developed innovative low-rate DoS attacks, like SlowITe and SlowTT that exploit these weaknesses. His research sheds light on the need for robust security measures in IoT environments to mitigate these vulnerabilities and ensure reliable operation.

Table 3 presents a comprehensive overview of the most impactful documents, offering an essential foundation for a bibliometric analysis.

These most cited studies explore various aspects of privacy and security in emerging technologies, from wireless networks to the IoT and even medical applications. Capkun and Hubaux (2006) address vulnerabilities in positioning techniques to spoofing attacks, proposing a mechanism called “verifiable multilateration” to secure location data in wireless networks. Chaabouni et al. (2019) delve into the security challenges of IoT networks, analyzing existing intrusion detection systems and advocating for ML techniques to enhance detection capabilities. Kalnis et al. (2007) tackle the issue of privacy in location-based services, proposing methods for anonymizing spatial queries while preserving result accuracy. Song et al. (2020) introduce a framework that leverages blockchain and deep learning to improve vehicular positioning accuracy and security in the Internet of Vehicles. Li et al. (2020) suggest a blockchain-based system for secure and efficient global positioning system (GPS) error sharing among vehicles, contributing to more accurate positioning in autonomous driving. Finally, Kumar et al. (2021) explore the use of blockchain and federated learning in medical applications, presenting a framework for collaborative training of deep learning models for COVID-19 detection and ensuring data privacy.

The co-citation network analysis in this study employs a visual representation to illustrate the interconnectedness of different research articles based on their shared citations. This network highlights which papers are frequently cited together, indicating their potential conceptual or methodological connections. This analysis helps identify clusters of research themes and understand how knowledge within a field evolves (Zhang et al., 2021). Figure 2 illustrates the references co-citation network. A threshold of seven citations has been used so that only references that have been cited at least seven times in the study corpus will be represented in the network map. Out of 4,747 references cited, only 115 reach this minimum threshold, which makes it possible to filter and focus only on the most influential and frequently referenced works in the field. It identifies three distinct clusters, represented in red, green, and blue, respectively, titled IoT devices, privacy-preserving techniques, and neural network architectures. Together, these clusters offer a comprehensive view of the technological innovations, ethical imperatives, and strategic opportunities that drive the evolution of MT marketing, ensuring that practices remain both cutting-edge and responsible.

Figure 2

References co-citation network.

This cluster provides a valuable overview of the research surrounding the proliferation of IoT devices. It highlights the rapid growth of this technology and its potential to revolutionize various industries and aspects of our lives. The studies by Al-Fuqaha et al. (2015) provide a foundational overview of IoT technologies and applications, setting the stage for subsequent research. Raza et al. (2013) address early security concerns by introducing SVELTE, an intrusion detection system for resource-constrained IoT devices. As threats evolved, Meidan et al. (2018) proposed N-BaIoT, an advanced network-based anomaly detection method using deep autoencoders to identify compromised IoT devices. To support ongoing research and development, Koroniotis et al. (2019) developed the Bot-IoT dataset, incorporating legitimate and simulated IoT network traffic with various attack types.

This cluster encompasses studies that showcase significant advancements in privacy-preserving techniques across diverse data types and applications. Dwork et al. (2006) introduced the concept of differential privacy, proposing a method to add calibrated noise to statistical database queries to preserve individual privacy while maintaining data utility. Gruteser and Grunwald (2003) developed a middleware architecture for location-based services that adjusts the spatial and temporal resolution of location data to meet anonymity constraints. Sweeney (2002) presented the k-anonymity model for protecting privacy in released datasets, ensuring that each individual’s information is indistinguishable from at least k − 1 others. Finally, Beresford and Stajano (2004) introduced the concept of mix zones for location privacy in pervasive computing, along with metrics for assessing user anonymity in frequently changing pseudonyms.

The studies in this cluster highlight significant advancements in neural network architectures and datasets, driving progress in the field of computer vision. Krizhevsky et al. (2017) introduced AlexNet, a breakthrough convolutional neural network that significantly improved ImageNet classification. Moreover, He et al. (2016) presented ResNet, introducing residual learning for training very deep networks. Simonyan and Zisserman (2014) proposed VGGNet, demonstrating the importance of network depth with small convolutional filters. Finally, Deng et al. (2009) introduced ImageNet, a large-scale image database that became crucial for training and evaluating deep learning models.

The thematic map in bibliometric analysis organizes research themes based on centrality (importance) and density (development level) (Callon et al., 1991; Cobo et al., 2011; Meyer et al., 2023). This visualization helps researchers identify research gaps, emerging trends, and the overall landscape of a field, highlighting potential areas for investigation. Figure 3 presents the thematic map for this study. The parameters used to construct this thematic map include a selection of the 250 most frequent words, with a minimum threshold of five occurrences per thousand documents for the formation of clusters. Each cluster is limited to a maximum of three labels with a size of 0.3 to optimize readability. The Louvain algorithm has been applied for the detection of thematic communities, while a community repulsion parameter of 0.1 ensures an appropriate visual separation between the different groups of concepts on the map.

Figure 3

Thematic map.

The conceptual maps in this section are derived from keyword clusters, with their visual representations generated using the PlantUML website. These maps are constructed based on the frequency and co-occurrence of keywords within the literature.

This cluster examines the transformative impact of AI and ML on marketing practices, focusing on customer segmentation, predictive analytics, and customer behavior analysis. The studies reveal how AI is being used to optimize audience targeting by going beyond traditional optimal distinctiveness and tailoring strategies to different revenue models (Majzoubi & Zhao, 2023; Van Angeren et al., 2022). Predictive models are being implemented to remedy crucial challenges like food security, crop yield estimation, and financial risk assessment (Gambetti et al., 2022). AI is additionally empowering personalized recommendations and driving strategic decision-making in areas like advertising and IoT adoption (Shi et al., 2022). However, the review highlights the need to address privacy concerns, ensure explainability, and foster continuous learning to promote accountable and effective AI deployment in the ever-evolving marketing landscape.

Clark et al. (2023) address the issue of “user disambiguation,” which aims to identify individual viewership within multi-person households. Their ML approach provides accurate insights for television ad targeting and audience analysis, a valuable tool for marketers in the entertainment industry. Additionally, Peng and Xin (2019) propose a social trust and preference segmentation-based recommendation algorithm to improve personalized suggestions. By incorporating trust relationships and preference domains, this algorithm enhances the effectiveness of recommendation systems in e-commerce platforms. Furthermore, an AI-based real-time positioning advertising system leveraging 5G base station data enhances the efficiency of audience targeting and personalized advertisement delivery, unlocking new opportunities for location-based marketing strategies (Cheng & Wang, 2021).

This cluster examines the complex relationship between privacy technologies, data handling, learning and modelling, positioning and navigation, advanced technologies, and finally ethics and safety.

The explosive growth of data, particularly location data from GPS-enabled devices, offers unprecedented opportunities for personalized marketing. AI-powered algorithms perform accurate analyses of these data to predict consumer behavior, segment customers, and tailor marketing messages (Benslama & Jallouli, 2020, 2022, 2024; De Sousa et al., 2023). However, this level of personalization comes at a cost: consumer privacy. The studies reviewed highlight several key concerns and trends related to positioning and navigation, advanced technologies, ethical and safety-privacy technologies, data handling learning and modelling. Results show, for example, that location data from GPS-enabled devices, such as smartphones, are incredibly valuable for understanding consumer behavior and creating highly personalized marketing campaigns (Nanni et al., 2016; Stachl et al., 2020). However, location data are also highly sensitive and can reveal a vast amount of personal information (Michael & Clarke, 2013; Simpson, 2014). Consequently, there is an urgent need for robust privacy-preserving techniques to anonymize or obfuscate location data before it is used for marketing purposes (Kalnis et al., 2007; Peng et al., 2014). Studies demonstrate that even anonymized location data can be re-identified, requiring advanced privacy-enhancing solutions (Hoh et al., 2010).

As a result, consumers are increasingly aware of and concerned about the collection and use of their personal data, particularly location data (Banerjee, 2019; Milne & Bahl, 2010). This awareness is leading to a demand for more transparency and control over data usage (Yang & Kels, 2016). Public trust in data service providers varies depending on their familiarity and the type of data collected (Banerjee, 2019). Therefore, there is an expanding need for ethical considerations and regulatory frameworks to address these concerns (Milne & Bahl, 2010; Teodorescu et al. 2025; Yang & Kels, 2016). Federated learning is emerging as a promising solution for balancing the needs for personalization and privacy (Cheng et al., 2022; Huo et al., 2023). Several studies showcase the potential of federated learning for privacy-preserving marketing applications, including user profile construction (Huo et al., 2023) and travel time prediction (Zhu et al., 2021).

The niche cluster of sensors identifies a critical research gap in the literature concerning the integration of sensors, global navigation satellite system, and various data security techniques within emerging technologies such as smart cities and cyber-physical systems. Zakroum et al. (2022) and Orouji and Mosavi (2020) demonstrate the vulnerabilities of network telescopes and IoT devices to attacks, emphasizing the need for robust security measures. Simultaneously, research on privacy-preserving techniques, such as differential privacy and anonymization, explored by Liu et al. (2023), remains critical to protecting user data in increasingly sensor-rich environments. Researchers like Kong et al. (2022) and Wu (2022) are tackling these challenges by incorporating federated learning and blockchain technologies within 6G networks and intelligent transportation systems. However, the integration of sensor phenomena and characterization, along with the complex dynamics of cyber-physical systems, requires further investigation to ensure responsible and secure deployment of these advanced technologies.

Results illustrate the multifaceted nature of security, structured around six main dimensions: Data Security, Enabling Technologies, Marketing Security, Trust and Reputation, MT Security, and Network Security. Data Security encompasses privacy measures, Encryption, and Cybersecurity to protect sensitive information. Enabling Technologies include blockchain, 5G, cyber-security, and authentication to enhance security across various platforms. Results show that Marketing security was studied mainly concerning the concept of cyber-security. The branch of Trust and reputation focuses on the authentication concerns. MT Security focuses on IoT, device protection, and GPS security. Finally, Network Security involves intrusion and anomaly detection to safeguard against unauthorized access and cyber-attacks.