Authors - Reena (Mahapatra) Lenka, Jaee Jogalekar Abstract - The outward appearance is a crucial necessity for all organizations and sectors. Daily attendance registration in a conservative manner is a tedious and lengthy task. Furthermore, each organization has sanctioned its own approach to sparkle involvement through personal mockery and the employment of sheets to bolster its attendance. To tackle these challenges, various standard automated identification and verification systems have been extensively utilized, including IRIS, RFID, and biometric techniques. In contrast, crocodile growth is highly demanded under these conditions, requiring more time, and it is reckless in vegetation. Any error or harm to the RFID card will lead to incomplete participation Locating this array of strategies for such a broad range necessitates higher costs and additional effort to integrate our involvement into the database. In today's context, the awareness and recognition of unique identities have grown globally due to the demand for safety in financial transactions, health monitoring, validation, and security, along with other essential factors such as reducing fraudulent participation, increased costs, and, most importantly, lowering the chances of marking our involvement. This document outlines a suggested enhancement for the clever attendance verification system, incorporating a facial recognition approach utilizing deep learning techniques. A cloud dataset will mainly be created by capturing the faces of the approved students or staff. Similarly, the face is affirmed through the inclination derived from deep learning. In addition, the created images will be saved in the established database, each assigned a distinct label. The extraction of facial features will rely on Haar-like characteristics computed through a Deep Learning method. The suggested new method attains improved propagation by utilizing Haar-like features and a deep-learning-optimised algorithm
Authors - Alta van der Merwe, Mpho Xaba Abstract - This paper examines how change management can be integrated into the TOGAF Architecture Development Method to strengthen enterprise architecture implementation and organisational transformation. Using a qualitative systematic literature review of 35 studies published after 2010, we identified recur-ring change management dimensions across models, theories, frameworks, and methodologies, namely leadership, strategy, communication, organisational structure, organisational culture, collaboration, transformation, innovation, governance, risk management, and commitment. We then aligned these dimensions to relevant TOGAF ADM phases and illustrated their application through a fictitious case study of Teleconnect, a telecommunications company undergoing post-acquisition integration. The findings show that TOGAF implementation is strengthened when change management is embedded as a structured and complementary organisational capability rather than treated as a separate activity. The paper contributes a conceptual framework that links change management dimensions to TOGAF ADM and offers a practical basis for supporting enterprise trans-formation through a more integrated architecture approach.
Authors - Alta van der Merwe, David Moselane Abstract - This study examines how artificial intelligence can strengthen teaching practices and improve university readiness for first-year students within the context of Society 5.0. While AI offers strong potential to support a human-centred and inclusive education system, its implementation faces obstacles, including resistance to change and scepticism about its value. The research explores strategies for effective AI integration, with a focus on personalised learning experiences tailored to individual student needs and the automation of administrative tasks to allow educators to focus on improving their teaching and student engagement. A systematic literature review and meta-analysis were conducted to evaluate how AI enhances university preparedness, with particular attention to perceptions of AI adoption and the challenges associated with its implementation. The findings highlight how AI can support more inclusive and responsive education systems aligned with the goals of Society 5.0, where technology serves societal needs. While acknowledging limitations such as time constraints and the rapidly evolving nature of AI technologies, the study offers practical insights to help educators reduce educational disparities, promote inclusivity, and equip students with skills required for a socially responsive and technologically integrated.
Authors - Franck W. Boubayi, Regis F. Babindamana, Peter A. Kidoudou Abstract - The adoption of e-commerce remains a major challenge for many enterprises in developing countries, where digital transformation is often constrained by technological, organizational, and regulatory factors. This study investigates the factors influencing e-commerce adoption among Congolese enterprises through the Technology Acceptance Model (TAM). Data were collected from businesses operating in various sectors and analyzed using descriptive statistics and predictive modeling techniques. The findings reveal that although digital technologies are increasingly used in business activities, e-commerce adoption remains limited. Only 35.5% of surveyed enterprises actively engage in online sales, while most organizations continue to rely on social media platforms for promotion and telephone-based order processing. The results also highlight significant cybersecurity gaps: nearly half of the surveyed enterprises do not conduct vulnerability assessments, more than half lack firewall or intrusion detection mechanisms, and 16% have already experienced cyberattacks. In addition, limited awareness of national data protection regulations exposes many businesses to legal and operational risks. The study further shows that artificial intelligence is widely perceived as a strategic opportunity, with 83% of respondents recognizing its potential for business growth and 89% supporting the establishment of an appropriate regulatory framework. Based on the identified determinants, a predictive model is proposed to support decision-making and promote wider adoption of e-commerce in the Congolese context. The findings provide practical insights for policymakers, business leaders, and researchers seeking to accelerate digital transformation while strengthening cybersecurity readiness.
Authors - Oleksandr Hladkyi, Alexander Gertsiy, Tetiana Tkachenko, Valentyna Zhuchenko, Tetiana Shparaga, Olha Liubitseva, Tetiana Mykhailenko, Iryna Kochetkova Abstract - The profitable spatial location of tourism companies and destinations plays an important role in land management investigations nowadays. It significantly influences on tourism destinations' spatial efficiency. There are four main concepts of determining spatial efficiency of enterprises' location: the urban planning concepts, synergistic or integrative concept, functional and communicative concept as well as the concept of service clusters. Our approach is essentially different from all mentioned above. It's based on the analysis of tourism products (goods, labor, services) production efficiency rates determined by spatial location effect in land management system. The process of estimation of spatial efficiency of tourism destinations' location should be divided into four parts. The first part consists in gathering complete statistical data about tourism destinations development in specific location/region. Based on primary statistical data, the following indicators of tourism destinations' spatial efficiency could be calculated: labor productivity, profitability, capital-labor ratio and cost recovery. At the second part, every index of tourism destinations' spatial efficiency has to be modulated using gravitational potential model. At the third part we have to identify individual clusters of different levels of tourism destinations' spatial efficiency based on gravitational modulator data. At the fourth part all received clusters could be figured on geographic contour maps of particular region. Using above-mentioned methods, gravitational model of spatial efficiency of tourist enterprises' location in key regions of Ukraine has been created.
Authors - Franck W. Boubayi Abstract - The rapid development of e-commerce platforms has significantly increased the exposure of digital systems to advanced cyber threats such as phishing, DDoS attacks, identity theft, and data breaches. To address the limitations of conventional security mechanisms, this paper proposes a Risk-Adaptive Multi-layer Security Framework (RAMSF) integrating multi-factor authentication, AI-based intrusion detection, post-quantum cryptography, and blockchain-based distributed auditing. The main contribution of the model lies in an adaptive decision engine based on dynamic risk evaluation, enabling real-time adjustment of security policies according to behavioral context and detected anomalies. The framework is evaluated using the CICIDS2017 and UNSW-NB15 datasets with 10-fold cross-validation. Experimental results show that RAMSF outperforms several classical models, including SVM, Random Forest, CNN, and XGBoost, achieving 97% accuracy, 95.5% F1-score, 98% AUC, and a low false positive rate. These results demonstrate that adaptive hybrid architectures represent a promising approach for strengthening cybersecurity in e-commerce systems against both current and post-quantum threats.
