Authors - Polinati Vinod Babu , M.V.P Chandra Sekhara Rao , Kolagotla Venkateswara Reddy , Manukonda Ravi Chandra , C P Pavan Kumar Hota , Kurumalla Suresh Abstract - The web services today are flooded with automated bots, distributed request floods and IP based attacks. Traditional firewalls are based on either a static signature rule or an opaque machine learning model that is not very transparent. We introduce ShieldNet, a lightweight Web Application Firewall (WAF), which is developed in FastAPI and Redis as a state storage. The explainable real-time request filtering performed by ShieldNet is on the basis of a deterministic decision pipeline, a combination of IP reputation scoring (Redis backed, counts on infractions), SlowAPI based rate limiting, Detection of bot user agent, Up-to-date use of open access threat feeds (spamhaus, abuse.ch, etc.). Request identities can be provided as either permitted (HTTP 200), rate limited (429) or blocked (403); all of which can be traced through the middleware layers. Synthetic load testing and feed validation reveal low latency, high throughput and high detection fidelity. The system has good performance, scalability and interpretability balance thus it can be used in real-time web security implementation.
Authors - Apolinar P. Datu, Pamela V. Zuniga, Chona S. Lajom, Jobert D. Bravo, Renen Paul M. Viado, Ma. Yvonne Czarina C. Angcaya, Maricris Punzalan Abstract - Higher education institutions play a critical role in advancing sustainable development, particularly through effective and inclusive communication practices supported by digital technologies. This study examined the role of Information and Communication Technology (ICT) standards in shaping effective communication strategies for sustainable development in higher education. Using a quantitative descriptive research design, data were collected through a structured questionnaire administered to thirty (30) higher education personnel, including faculty members, administrators, and ICT staff aged 22 years and above. A 4-point Likert scale was utilized to measure levels of ICT standards implementation, communication effectiveness, accessibility and inclusivity, system interoperability, and stakeholder engagement. The findings revealed a high level of ICT standards implementation across participating institutions. Results further showed that ICT standards strongly influence the clarity, reliability, timeliness, and organization of communication strategies. Standardized ICT practices were also found to enhance accessibility and inclusivity in digital communication, support system interoperability, and improve stakeholder engagement in sustainability-related initiatives. Overall, the study highlights that ICT standards serve as essential enablers of coherent, inclusive, and sustainable communication in higher education. The findings underscore the importance of strengthening ICT standards to support institutional sustainability goals and foster meaningful stakeholder participation.
Authors - A Aruna kumari, Sri Vishnu Prabhu Gudavalli, Tamminana Visweswari Abstract - This project presents a hybrid architecture, which combines sequential and non-sequential models, to detect diabetic retinopathy (DR) in retinal fundus image data using deep learning. The model has a non-sequential backbone feature extraction layer, a pre-trained ResNet50, built using the Functional API of Keras to allow flexibility and skip connections. On top of that, a functional custom sequential classifier head is added, comprising of, e.g., GlobalAveragePooling2D, Dense, and Dropout layers, also constructed in a functional manner to ensure the architecture is coherent. The classifier is used to make a binary prediction: DR or not. The model was trained using the EyePACS data using different techniques of image preprocessing, including resizing, normalization, and augmentation. The resulting architecture is completely accurate and exhibits good generalization on unseen data, which indicates the effectiveness of transfer learning in combination with deep sequential classifiers. This bifurcated architecture is understandable and performance-wise attractive to be used in clinical decision support in diabetic retinopathy screening.
Authors - Masaki Murakami, Atsuhiro Goto Abstract - Post-quantum cryptography (PQC) migration in the financial sector is a resource allocation problem shaped by network interdependence. This study develops a prototype dynamic simulation model that incrementally introduces cyber-risk internalization: starting from a baseline without cyber-risk channels, adding firm-level direct cyber loss, and finally incorporating systemic cyber loss propagated through the financial network. The paper quantifies how the scope of cyber-risk internalization determines migration outcomes under a shared investmentallocation framework. Results show that firm-level internalization accelerates early migration but remains insufficient for full sector-wide coverage, whereas systemic-risk internalization can close the late-stage adoption gap under sufficiently supportive policy conditions. This distinction shows that policy support for PQC migration should not be understood only as cost reduction, but also as a mechanism for internalizing network externalities. Through nested comparisons of these model variants, this study demonstrates how each risk channel distinctly shifts migration outcomes.
Authors - Akbar Doniyorovich, Doniyor Gulomov Zaynobidin o'g'li, Dilshoda Akramova, Xushmamatova Aminakhon Rustam qizi, Tukhtaeva Shakhnoza Gaybulla kizi, Abdimurodova Shakhnoza Anvar qizi Abstract - This article discusses the theoretical foundations and рractical directions of innova tive didactic methods aimed at develoрing the comрetence of using artificial intelligence (AI) in рedagogical activities. The study analyzes the рroblem of forming teachers' digital literacy, technological readiness, and ability to integrate intellectual technologies into the educational рrocess. The article рroрoses methods aimed at increasing the innovative comрetence of teach ers based on a didactic aррroach and a steр-by-steр learning model develoрed to achieve effec tive results in the use of AI tools in education. These methods include aррroaches such as рroblem-based learning, рroject work, reflexive analysis, and collaborative learning with AI as sistants. According to the results of the study, the systematic introduction of artificial intelli gence technologies into the educational рrocess develoрs the information-analytical, methodo logical, creative, and reflexive comрetence of teachers. The use of innovative didactic methods increases the effectiveness of teacher activity, individualizes the educational рrocess, enhances analytical thinking, and a creative aррroach. The scientific results рresented in the article serve as a scientific and methodological basis for the рrocess of modernizing the рedagogical education system, imрlementing the conceрt of digital рedagogy into рractice, and рreрaring future teachers to work with artificial intelligence technologies.
Authors - Apolinar P. Datu, Jobert D. Bravo, Reine Joshua L. Cruz, Helga Marie B. Cabarle, Najera R. Umpar, Minsoware S. Bacolod Abstract - In today’s rapidly evolving Information and Communication Technology (ICT) landscape, digital tools have become essential in shaping how research is presented, shared, and collaboratively developed. This study aimed to quantitatively examine the collaborative benefits of digital tools and their impact on user engagement during research presentations. Using a quantitative descriptive–comparative research design, data were collected from thirty (30) respondents consisting of students, educators, ICT professionals, and administrative staff who regularly engage in ICT-based collaborative activities. A structured survey questionnaire was used to gather numerical data on the frequency of tool usage, levels of collaboration, and user engagement. Descriptive statistics such as frequency, percentage, and weighted mean were employed to analyze the data. The findings revealed that digital collaboration tools are frequently used in ICT-related activities and are strongly associated with improved coordination, communication, and teamwork. Respondents reported high levels of collaboration and engagement, particularly when using interactive features such as real-time communication, shared document editing, and feedback mechanisms. These features were found to enhance motivation, focus, and willingness to participate actively in research-related tasks. Overall, the study highlights the importance of intentional and effective integration of digital tools in research presentations to foster meaningful collaboration and sustained engagement in ICT-based academic and professional environments.
Authors - Khumbelo Difference Muthavhine, Mbuyu Sumbwanyambe Abstract - Inadequate Resource Issues (IRI) are one of the challenges in Strategic Management (SM). This study concentrated on applying System Dynamics (SD) modeling to solve IRI. Strategic standard tools like SWOT analysis, PESTEL analysis, and the Resource-Based View have proven effective in addressing IRI; unfortunately, developments like digital transformation, long-term sustainability, and the rise of emerging market multinational corporations are poised to shape the future of SM in these regions. These traditional methods are no longer coping with new technology; hence, the authors implemented a new SD model to tackle the IRI in SM. Additionally, most strategic managers are incapable of developing an SD model due to mathematical and scientific complexity. Although SD is a reliable technique for handling complex issues in management, most managers reject SD because of the implementation’s need for scientific and mathematical requirements. To solve IRI mathematically and make scientific predictions about what would happen if variables were altered in the upcoming five years (2025–2035) and the impact on customers, the study created an SD model.
Authors - Arosha de Silva Abstract - The COVID-19 pandemic disrupted educational systems worldwide and required schools to adopt online learning within a short period. In Sri Lanka, secondary school teachers encountered numerous difficulties while adapting to virtual teaching environments. This study examines the challenges experienced by teachers when conducting online instruction during the pandemic. A mixed-methods approach was adopted, combining qualitative interviews with quantitative survey data collected from secondary school teachers and educational professionals. The findings revealed that teacher motivation, technological infrastructure, and increased workload significantly influenced the effectiveness of online teaching. Difficulties related to internet access, digital resources, and professional demands affected teachers’ ability to deliver lessons efficiently. The study highlights the importance of institutional support, professional training, and improved access to technology in strengthening online education. The findings may assist policymakers and educational institutions in developing effective strategies to support online and blended learning initiatives in the future.
Authors - Bharg Trivedi, Chaitaili Chandankhede Abstract - There is a continuous change in Android malware because it is obfuscated, polymorphic, and structural. Such changing methods diminish the performance of conventional signature-based detection methods. In an effort to defeat this challenge, the present paper provides a model that uses CNN and GNN models. It is an integration of spatial byteplot representations and structural call graph representations to successfully identify Android malware. Our study was based on a dataset of 1,159 real Android applications, and the used extraction technique was based on the static features. The CNN element of the structure Recognized robust spatial attributes of the grayscale images of the byteplot data with a ResNet-50 network. Meanwhile, the GNN component of the structure used a GraphSAGE network to derive structural representations of automatically generated function call graphs. The fused representations are combined into a 2304 dimensional feature vector. It is also optimized by making use of different methods such as Mutual Information. In this study, an Extreme Gradient Boosting Classifier on the fused representations to achieve successful Android malware detection. The assessment indicates that the framework attains a classification accuracy of more than 99% with businesses across the cross-validation holding the same accuracy.
Authors - Khumbelo Difference Muthavhine, Mbuyu Sumbwanyambe Abstract - Knowledge management (KM) is an essential company training process that incorporates four sequential factors: non-knowledgeable professionals, training to become knowledgeable professionals, new knowledgeable professionals, and knowledgeable and experienced professionals. Because the aforementioned variables are interconnected and make it extremely difficult to produce a measured solution, they must be thoroughly analyzed using mathematical formulas and reliable techniques. These issues impact businesses of all sizes, necessitating a versatile instrument for flexible KM analysis. Additionally, most KM managers dislike SD modeling due to its complexity, especially those without scientific training. This study recommended using system dynamic (SD) modeling rather than conventional tools to address the aforementioned issues. The use of SD modeling stems from three factors: (a) the examination of complicated dependencies; (b) the requirement for mathematical formulas; and (c) the graphical results in contrast to traditional methods. The study’s SD model included the four sequential factors and their relationships. KM managers should focus especially on the graph’s data when necessary modifications are needed.
Authors - A Aruna kumari, Tamminana Visweswari Abstract - Lung cancer is a fatal illness that causes several deaths worldwide and detection of lung cancer remains a challenge for medical professionals. Detection of cancer in early stages is difficult as the size of the tumor is very small making it difficult for medical professionals to detect. Cancer detected in the early stages can be treated with proper techniques which can save the lives of the patients. Due to excessive information in the CT scans, MRIs, X-rays, and PET scans the manual detection of lung tumor becomes extremely difficult. The methodology helps in detecting the presence of cancerous tissues in the lungs and predicting which stage of lung cancer is present. The methodology mainly includes image preprocessing, training the model, extracting features using deep learning algorithms and classifying the stage of cancer present as Normal, Benign, Malignant Stage 1, Malignant Stage 2, and Malignant Stage 3. Proper image processing techniques like image augmentation, image normalization and image resizing are applied on the IQ-OTH/NCCD dataset for extracting the necessary features which will be used while training the model. A hybrid model is created by combining two deep learning models, the Xception and MobileNetV2 architectures which can accurately distinguish between the different lung cancer stages and predict the stage of cancer. The performance metrices which include accuracy, precision, recall, f1-score and confusion matrix were also calculated to determine the accuracy of the proposed hybrid model. The proposed model helps in accurate and reliable diagnosis of lung cancer at early stages.
Authors - Erika Haydee Rubio-Camara, Oscar May Tzuc, Elsy Maria Rosales-Uc, Fran-cisco Gilberto Herrera-Chale, Roman A. Canul-Turriza, M. Jimenez Torres Abstract - Mechanical vibration energy harvesting has emerged as a promising strategy for supporting sustainable energy generation in industrial environments, where machinery and transportation systems continuously produce recoverable vibrational energy. This study presents the development and evaluation of predictive models based on deep multilayer perceptrons (DMLP) and Convolutional Neural Networks (CNNs) for estimating the energy potential associated with mechanical vibrations under industrial operating conditions. A simulation frame-work was implemented using experimentally reported operational ranges, including vibration frequencies between 10 and 50 Hz, amplitudes from 0.01 to 0.03 m, and temperatures between 25 and 45 °C. The analysis considered piezoelectric, electromagnetic, and triboelectric harvesting mechanisms to evaluate model adaptability under different scenarios. The predictive framework was implemented using TensorFlow and validated through a 10-fold cross-validation strategy combined with hyperparameter optimization. Results indicate that both architectures achieve high predictive capability for estimating harvested energy; however, CNN models consistently outperformed Deep MLP models, obtaining lower prediction errors and higher stability across validation folds. The superior performance of CNNs is associated with their ability to capture localized patterns and structured relationships within vibration-related data. The proposed method-ology demonstrates the feasibility of integrating artificial intelligence techniques into vibration-based energy harvesting systems for industrial applications. Furthermore, the study provides a computational framework for evaluating operational conditions, optimizing harvesting performance, and supporting the design of sustainable self-powered monitoring systems.
Authors - Khumbelo Difference Muthavhine, Mbuyu Sumbwanyambe Abstract - The financial management involves outstanding debts, trade receivables, deduction charges, economic value, generating capacity, and stockpiling accumulation. These six variables mentioned above must be carefully examined using mathematical formulas and trustworthy tools because they are interrelated, which makes it very difficult to establish a solution. These challenges affect companies of all sizes, requiring a flexible tool for adaptable financial management analysis. To mitigate the aforementioned problems, this study suggested system dynamic (SD) modeling to handle the problem instead of using traditional tools. The SD modeling is used because of (a) complex dependencies analysis, (b) the need for mathematical formulas, and (c) the graphical outputs compared to traditional tools. The study constructed an SD model with six variables and their interconnections. When adjustments are required, financial managers should pay particular attention to the out-of-graph data.
Authors - Mariusz Szynkiewicz Abstract - The protection of information resources is one of the central issues in contemporary information science, and one also significant from an IT perspective – particularly in the context of cybersecurity. In this article, I propose one possible approach to addressing this challenge. The proposal concerns the protection of information resources in a broad sense: from the stage of information acquisition and creation through to its distribution. In the following sections, I outline the main assumptions of a comprehensive model for the protection of information resources, discussed in the context of building the information resilience of participants in digitised information exchange processes, and in relation to issues associated with the concept of cyber hygiene. The central thesis of the article rests on the assumption that effective protection of information resources is possible only on the basis of an integrated model addressing the following procedures: (a) validation – the assessment of the level and value of a given resource; (b) threat identification – the logical level, scale, and types of vulnerability; (c) detailed analysis of the type of abuse – qualitative diagnosis; (d) selection of techniques and methods for counteracting a given class of attack – the methodological level; and (e) selection of possible corrective and preventive measures – the elements of cyber hygiene.
Authors - S. Jimenez-Garcia, V. Zorrilla-Munoz, G. Martinez-Navarrete, N. Garcia-Aracil, A.M. Peiro-Peiro Abstract - This paper presents an integrated framework that combines digital screening, ergonomic assessment, humanoid robot benchmarking and immersive virtual reality (VR) training to support the prevention of musculoskeletal disorders and burnout in nursing professionals. The framework is grounded in occupational health data from Spanish nursing professionals and incorporates sex/gender and anthropometric differences as design variables. A descriptive cross-sectional analysis was performed on 316 professionals from the European Health Survey in Spain, filtered by occupation, and complemented with the PROBEREN project approach. Two high-demand activities in Internal Medicine and Infectious Diseases units were selected: hygiene, comfort care, pressure ulcer prevention and postural changes in bedridden patients; and intrahospital transfers under isolation or clinical support. The sample showed a strong proportion of women (86.4%), mean age of 45.97 years, chronic health problems in 54.7%, prescribed medication use in 51.9%, recent physical pain in 46.8% and pain interfering with daily activities in 31.6%. These findings support a transition from descriptive profiling to proactive prevention. The proposed ecosystem links early screening, capture of expert movements, biomechanical comparison with a humanoid robot, personalized VR training and longitudinal reassessment. Future pilot validation should evaluate usability, VR-related fatigue, ergonomic risk reduction, burnout, pain and implementation barriers in real clinical settings.
Authors - Aleksander Karastoyanov Abstract - Reinforcement learning (RL) has been widely proposed for adaptive virtual machine (VM) right-sizing in cloud environments, yet most published work reports results from a single random seed with a fixed reward formulation, conditions that may not reflect genuine generalization. This paper addresses both limitations through a systematic multi-seed, multi-reward evaluation of a Proximal Policy Optimization (PPO) agent applied to VM right-sizing on the Alibaba Cluster Trace 2018. Fifteen independent training runs (three reward configurations × five seeds) are conducted on a ten-VM simulation environment. The key finding is that reward formulation, not the RL algorithm per se, is the dominant determinant of SLA compliance: the unified reward variant (Config A) achieves a mean SLA violation rate of 1.1% (±2.11) across five seeds, statistically comparable to a Threshold baseline (2.43%), while dimension-aware variants expose a critical instability in memory-saturated environments. A one-sample t-test yields t = −1.40, p = 0.23, confirming that neither superiority nor inferiority relative to Threshold can be claimed, and motivating the need for larger seed sets and alternative reward designs in future work. Three empirically grounded reward design principles are derived for practitioners deploying RL-based resource managers in high-memory-pressure infrastructure.
Authors - Soham Paithankar, Supriya Narad Abstract - Agriculture is a major contributor to India economy and supports the livelihood of a large population, however, traditional farming practices largely depend on manual observation, weather information, which often results in resource utilization and reduced crop productivity. The increasing variability of climate condition further challenges. The integration of internet of things (IoT) and software technologies provides an effective approach and data-driven decision by enabling real-time monitoring and data-driven decision making. This paper presents such as temperature, humidity, and soil moisture using field sensor data. The collected information is processed and stored using Python based software and compared with real-time weather information obtained through a weather API. A dashboard interface developed using Flask and streamlit visually presents sensor data, weather data, and comparative result to support informed decision-making system. The system demonstrates how low-cost IoT devices combine with software platform can improve agriculture monitoring, optimize resource usage, and support sustainable farming practice, this study highlights the potential of IoT and software integration in transforming the potential of IoT. agriculture into an intelligent data-driven system suitable for developing region such as India.
Authors - Emmanuel Opoku Debrah, Sunet Eybers, Corne J. van Staden Abstract - Student Information and Management Systems (SIMS) are increasingly being implemented in higher education institutions (HEIs) in developing countries. However, with these mandatory systems, there is limited evidence of user satisfaction. This systematic literature review investigates the influence of technological, organizational, and environmental (TOE) factors on user satisfaction with the use of mandatory SIMS in HEIs in developing countries, with a focus on Ghana. A search was conducted across six databases and Google Scholar, focusing on the past decade (2015-2025). Two reviewers independently evaluated the studies' quality using the Mixed Methods Appraisal Tool (MMAT 2018). The average consensus MMAT score was 4.08/5 (7 High, 14 Moderate, 5 Lower quality). A total of 1,382 records were screened and considered for duplicates and applicability. The result was 27 empirical studies. A narrative synthesis supported by thematic mapping identified the most frequently reported determinants: system quality, reliability, and performance (n=18); organizational support, IT capacity, and management readiness (n=12); perceived usefulness, ease of use, and user experience (n=10); ICT infrastructure and connectivity (n=9); and information quality (n=6). The synthesis suggests that TOE factors are not independent: technological advantages can be offset by organizational weaknesses, and environmental factors set an upper limit to satisfaction. This suggests that Ghanaian HEIs focus on integrated investments in user training, technical support, and infrastructure rather than just system upgrades.
Authors - Md Manirul Islam, Md. Mushfiqur Rahman , Sazzad Hossain Abstract - Moving IoT-edge systems to post-quantum security is not as simple as replacing one algorithm with another. Different parts of the system have different security needs. Short-lived telemetry, control messages, firmware updates, and trust records should not all be protected in the same way. Device limits are also very different: a tiny leaf sensor, a capable actuator, a gateway, and a cloud service do not have the same memory, energy, or bandwidth budget. This paper presents an adaptive hybrid classical-post-quantum cryptographic framework for such heterogeneous environments. We make four main contributions. First, we define a four-tier system and threat model for resource-constrained IoT-edge deployments. Second, we describe profile selection as a practical decision problem that balances security, latency, energy, memory, and bandwidth. Third, we propose a small profile catalog, P0 to P4, that maps classical, hybrid, and PQ-first choices to real deployment roles. Fourth, we evaluate the framework through a benchmark-grounded experimental emulation using current standards and published device measurements. The main idea is simple: apply the strongest protection where compromise would hurt most, while keeping weak devices usable in the real world.
Authors - Imane Bari, Mounir Oubenyahya, Abdellatif Aziki, Fouad Achemchem Abstract - Climatic conditions and water scarcity in the Souss Massa region of Morocco are causing a significant decline in fodder supply and threatening the continuity of dairy farming. In response to these structural constraints, dairy cooperatives have adopted hydroponic green fodder (HGF) cultivation, an approach that reduces water consumption and frees agricultural land for higher-value crops. This study analyses the financial performance and structural effects of HGF adoption in two dairy cooperatives selected as pilot cases. Using a descriptive and explanatory methodology based on accounting and financial records collected over 4 to 6 years, combined with a non-parametric econometric model, the study assesses production costs, financial viability indicators, and the effect of innovation on key financial ratios. Results confirm the financial viability of the HGF investment and document its temporary effects on financial independence and self-financing capacity, while highlighting the social dimension of this innovation in preserving livestock farmers’ livelihoods.
Authors - Swetha P. , Maniratnam, Prasad B Honnavalli Abstract - The dark web has been identified as a major source of organizational risk, facilitating underground markets for stolen credentials, proprietary data, and pre-attack threat intelligence. An organization lacking visibility in these underground marketplaces faces attacks entirely beyond conventional monitoring solutions. This paper introduces a realtime dark web monitoring system integrating anonymous browsing via Tor, Selenium WebDriver, and open-source Wazuh Security Information and Event Management (SIEM) to create a unified threat intelligence solution. The system performs continuous keyword searching for organizational name references and autonomously browses authenticated dark web marketplaces for content extraction. Identified events are serialized as structured JSON messages and ingested into Wazuh via custom decoder and rule definitions, providing actionable alerts on the analyst dashboard within seconds. The system has been evaluated over 24-hour continuous sessions on two live dark web markets, achieving 95% keyword detection accuracy, 1.4 seconds average alert latency to dashboard, and stable long-duration browser operation without application crashes. The system requires no commercial licensing, is fully configurable to organizational targets, and features native SIEM integration, making it a viable and accessible solution compared to proprietary dark web intelligence services.
Authors - Pam Cole Abstract - As artificial intelligence transitions from passive tools to autonomous agents, governance frameworks designed for slower systems are being asked to contain behavior they were not engineered to address. Two converging forces widen the resulting gap. Synthetic Trust describes the industrialization of unearned credibility, where AI systems simulate socio-emotional cues to suppress verification. Ungoverned Acceleration describes autonomous execution at a pace that outruns institutional oversight. Drawing on cross-industry research covering sixteen major AI systems, peer-reviewed findings on emotional manipulation, and documented incidents across sectors, this paper proposes the A.W.A.R.E. Governance Flywheel, organized around Identity, Trust, and Control spokes with Visibility, Attribution, and Containment as operational mechanisms, bound by Accountability. A comparative analysis of six frameworks across six continents shows no widely adopted framework currently operationalizes agentic governance at this specificity. The paper extends the framework with operational metrics, industry applications, adoption barriers, and three testable propositions.
Authors - Yasmine AGOUN, Cheikh SALMI, Nour El-Houda SENOUSSI Abstract - The rapid growth of the Internet of Things (IoT) has made cybersecurity prone to several vulnerabilities, highlighting the need for a semantic and well-organized structure for cybersecurity knowledge to ensure reliable threat detection and mitigation. In this paper, we propose IoTSecOnto, a largely automated pipeline for building a security-centric Internet of Things (IoT) ontology. The pipeline combines automated security literature mining, text analytics, natural language processing, Large Language Models (LLMs), and formal concept analysis. This approach reveals domain-specific concepts and relations and organizes them into a coherent hierarchy. A human-assisted review phase is needed to ensure the reliability and the accuracy of the derived security knowledge. In addition, the ontology is designed to be flexible and can be improved over time. IoTSecOnto is implemented using OWL 2, RDFLib, SPARQL, and SHACL constraints. We used a Mirai botnet and ontology quality metrics to demonstrate its effectiveness. The obtained results confirm the ability of IoTSecOnto to support knowledge sharing, automated reasoning, and improved threat analysis across diverse IoT settings.
Authors: Nellylyn Moyo, Naume Sonhera Abstract: The adoption of generative artificial intelligence (GenAI) has quickly evolved from its experimental stage to a question of institutional governance in higher education. Although initial approaches to the integration of GenAI in universities have been characterized by ad hoc trial-and-error methods and inconsistent policies, higher education organizations are now taking a more systematic approach to the regulation and implementation of GenAI for educational, research, and administrative purposes. This study provides a conceptual critical literature review of the approaches being adopted by universities toward GenAI, emphasizing trends in the global context and in Africa and South African higher education. The study demonstrates that GenAI readiness should not be viewed solely in terms of willingness to use AI technologies but rather as a multifaceted construct. Institutional governance, on the other hand, has evolved to incorporate responses related to issues of academic integrity, redesigning assessments, AI literacy, requirements for disclosure, development of staff, protection of data, and responsible-use guidelines. Nevertheless, such an evolution is patchy, reactive, and very much dependent on institutional capacities, disciplinary differences, and available resources. In the case of African higher education, the possibilities generated by the advent of GenAI have been heavily determined by structural limitations, such as digital inequality, uneven infrastructures, immature policies, and lack of capacity building measures. South Africa emerges as a particularly interesting example because of the emergence of governance responses in a space characterised by inequalities related to access, digital literacy, and institutional capacity. The study argues for an approach to GenAI adoption readiness that would consider the issue at three interdependent levels: individual readiness, organisational readiness, and structural readiness. Such an approach would enable a more contextually sensitive perspective on responsible GenAI adoption.
Authors: Nellylyn Moyo, Sello Prince Sekwatlakwatla, Tranos Zuva Abstract: GenAI has created considerable opportunities and challenges within institutions of higher education. While GenAI tools could help with teaching, learning, assessment, feedback, academic writing, research, and administrative functions, the uptake of these technologies cannot merely be viewed from a perspective of accessibility. In this study, an approach towards institutional readiness for adopting GenAI within higher education institutions is provided by examining infrastructure, governance, training, readiness for ethical adoption, reassessment, and fit-to-use practices. Peer-reviewed literature written between 2020 and 2026 is reviewed here to show that institutional readiness for GenAI adoption must be viewed as an organisational capability rather than as a static technology readiness state. The conclusion can be drawn that while reliable digital infrastructure is an indispensable requirement, institutional readiness also requires governance maturity, clarity of policy, integrity, protection of data, professional development of faculty members, AI literacy of students, as well as rethinking of assessments. The readiness concept is additionally influenced by factors such as digital inequality, varying capacities of institutions, limited resources, and policies in the higher education systems in Africa and South Africa. Fit to use alignment and practice-policy alignment are two concepts that may be useful in determining how well integrated the use of GenAI technology is into educational processes. GenAI can only be used responsibly if there is proper alignment among various aspects, including technology, policy, ethics, and context.
