The Personalized Internet of Things (PIoT) demands intelligent learning models that can adapt to highly heterogeneous user data while preserving privacy, scalability, and security. Centralized learning approaches are impractical in PIoT settings due to privacy regulations, non-independent and identically distributed (non-IID) data distributions, and vulnerability to adversarial attacks. To address these challenges, this paper proposes Consent-Driven Ethical Data Access and Regulation (CEDAR), a federated meta-learning framework that enables secure and adaptive personalization across the cloud–edge–device continuum. CEDAR integrates meta-learning with federated learning to extract transferable representations from distributed data, allowing rapid adaptation to individual user contexts with minimal local updates. A layer-wise adaptive uploading mechanism selectively communicates model updates based on parameter importance, substantially reducing communication overhead and accelerating convergence. In addition, asymmetric uploading and anomaly-aware aggregation enhance robustness against gradient inversion and model poisoning attacks. Extensive evaluations on six benchmark datasets covering regression, text classification, and image recognition tasks demonstrate that CEDAR achieves up to 60.39% higher accuracy compared to FedAvg-based federated learning, while reducing communication cost by 23.36% and improving adversarial robustness relative to other state-of-the-art baselines. Ablation studies further confirm the complementary contributions of CEDAR’s core components. By jointly optimizing personalization, privacy, efficiency, and security, CEDAR provides a scalable and ethically aligned learning framework for next-generation PIoT applications in domains such as smart mobility, healthcare, and the digital economy.

• View PDF

In the context of increasingly algorithmically driven digital platforms, cyberbullying has evolved into a complex communication phenomenon shaped by artificial intelligence, platform design, and automated content distribution. This study examines the prevalence and characteristics of cyberbullying among university students, focusing on awareness, reporting behaviour, and perceptions of institutional support within AI-mediated communication environments. The research was conducted on a sample of 67 university students using a structured online questionnaire. Results indicate that 30% of participants reported experiencing cyberbullying, while formal reporting to institutional authorities remained extremely low (3%). Although awareness of the term cyberbullying was high, only half of the respondents demonstrated a comprehensive understanding of the phenomenon. Anxiety, stress, and reduced self-confidence emerged as the most frequently reported consequences. The study further situates cyberbullying within contemporary developments in artificial intelligence, including algorithmic amplification and AI-supported content moderation, which influence the visibility of harmful content and user responses. Despite increased awareness, students rarely seek institutional support, often normalizing or ignoring abusive behaviour. The findings highlight the need for preventive strategies grounded in digital literacy, transparent AI governance, and science communication approaches that address both human and algorithmic actors, positioning cyberbullying as a critical challenge at the intersection of artificial intelligence, digital communication, and youth well-being.

• View PDF

As artificial intelligence (AI) systems become embedded in everyday communication, education, and decision-making, public AI literacy has become a pressing science communication challenge. This paper reconceptualizes AI literacy as a science communication task rather than a narrowly technical educational objective. Using a scoping-review-informed synthesis of interdisciplinary literature, the manuscript integrates scholarship from AI literacy, science communication, informal learning, and educational technology to derive the COMMUNICATE framework. The review drew on iterative searches of Google Scholar, Scopus-indexed sources available to the author, and backward reference tracing. Sources were included when they addressed AI literacy frameworks, public engagement and science communication theory, or empirical findings relevant to pedagogy, trust, participation, and AI-supported learning design. The resulting framework organizes eleven principles: Contextualized understanding, Open dialogue, Multimodal representation, Meaning-making, Universal accessibility, Narrative engagement, Interactive exploration, Critical evaluation, Adaptive scaffolding, Transformative learning, and Ethical reflection. The paper argues that a science communication perspective adds value by foregrounding audience diversity, public trust, dialogic participation, and interpretive context, which are often underdeveloped in technically oriented AI literacy models. The manuscript concludes by outlining practical implications for educators, communicators, and policymakers, while explicitly acknowledging the framework's conceptual status and the need for future empirical validation across formal and informal settings.

• View PDF