BOHR International Journal of Smart Computing and Information Technology

Anthropometric analysis and development of a simulink model or classifying body shapes using fuzzy logic

2025-12-11T10:10:46+00:00

This research analyzes the body shapes of 353 adult males (aged 18–60) in Southern Vietnam. The research employed a combination of cross-sectional statistical methods and fuzzy logic to ensure accuracy. Factor analysis first extracted four principal body dimensions. Using these as a basis, K-means clustering and analysis of variance (ANOVA) identified six distinct body shape groups. The author applied WHO standards, including body mass index (BMI), difference between chest and waist measurements (DROP), and waist-to-hip ratio (WHR), to differentiate between types, ranging from thin to obese. The average height of Southern Vietnamese men has increased compared to the TCVN 5782:2009 standard. This reflects a positive development in physical growth. The author developed a fuzzy logic model for rapid classification. The model uses height and weight as inputs to manage measurement uncertainty effectively. It has high practical value for the apparel industry, especially in 3D avatar generation. This research contributes to the scientific understanding of anthropometric variation. It also provides a foundation for personalized fashion and body shape prediction in Industry 4.0.

Adaptive context-aware middleware for seamless integration and service optimization in smart environments

2026-07-29T12:04:59+00:00

The rapid growth of the internet of things (IoT) and the rise of smart environments have increased the need for middleware systems that can support seamless communication among diverse devices while adapting to changing conditions. Many existing middleware solutions struggle with device heterogeneity, scalability, and real-time responsiveness, often leading to inefficiencies in performance and service delivery. To address these limitations, this study proposes an Adaptive Context-Aware Middleware framework designed to enable seamless integration and service optimization in smart environments. The framework continuously observes environmental and user contexts and dynamically adjusts system behavior to meet changing demands. It is built on four core components: context acquisition, context modeling, context reasoning, and adaptive management. These components work together to interpret real-time data and support intelligent decision-making within the system. The proposed framework was evaluated through simulation-based experiments under different workload scenarios. The results show that it consistently outperforms traditional static and conventional dynamic middleware approaches. In particular, it achieves lower latency, higher throughput, and improved system stability under increasing load conditions. The adaptive design also demonstrates strong scalability and flexibility, allowing it to operate effectively in diverse smart environments. These include applications such as smart homes, healthcare monitoring systems, intelligent transportation systems, and industrial IoT infrastructures. The findings indicate that adaptive context-aware middleware significantly improves system performance by enabling better integration of devices, optimizing resource usage, and enhancing service delivery. This makes it a promising approach for managing the complexity and dynamism of modern IoT-enabled environments.

Data mining in education: a comprehensive review

2026-08-05T11:26:08+00:00

Schools and colleges collect vast amounts of data from learning management systems, online learning platforms, assessment systems, and administration databases. Educational Data Mining (EDM) is a newly developing research area that is associated with the application of data mining techniques to educational data, the extraction of useful information from it, and the use of the information to enhance both instruction and learning as well as the decision-making processes of an educational institution. This review paper aims to present a comprehensive survey of the area of data mining applications, techniques, problems, and future directions in the educational sector. The current research focused on secondary data analysis of past studies grounded in the EDM and learning analytics (LA) field. Sources such as various published papers, journal articles, and conference papers were reviewed to find commonly used techniques, applications, and new trends in the field. We review some techniques like classification, clustering, association rule mining, predictive modeling, and LA. The results show that the role of this EDM in predicting students’ learning abilities, dropout rate detection, recommendation systems for student learning, and tools in curriculum designing has an important role. Moreover, data mining is used in the field of intelligent tutoring systems and in the analysis of the performance of institutions. The paper also identifies some significant issues in the field of EDM, such as data quality, privacy and ethical issues, and the requirement for understandable and accurate algorithms. In addition, new areas of research and development in educational technology are identified, including real-time analytics, multimodal data analysis, artificial intelligence (AI), and explainable AI. To summarize, by converting raw educational data into insightful information, EDM can be helpful to enhance the teaching, learning, and decision-making processes in educational activities and policies. The study underscores the need to use ethical and safe education data that could continue in the future for educational development while, at the same time, using effective data mining techniques.