Research Publications

1. Web Technologies Security in the AI Era: A Survey of CDN-Enhanced Defenses

Authors: Hosain, M.; Shuvo, S. A.; Ogbe, M.; Mazumder, M. S. J.; Rahman, Y.; Hakim, M. A.; Pandey, A.
Conference: IEEE Asia Pacific Conference on Wireless and Mobile (APWiMob 2025) – Emerging Services, Security, and Applications for Mobile & Wireless. (Accepted)

Summary:
This paper presents a comprehensive review of how Content Delivery Networks (CDNs) are reshaping the security landscape of modern web systems. With the rise of AI-driven attacks, CDNs now act as a critical defense layer for data-intensive and high-traffic web applications. The study examines current methods for traffic filtering, bot detection, and distributed denial-of-service (DDoS) mitigation. It also explores how machine learning and real-time analytics enhance CDN efficiency. The work concludes that integrating CDN-based defense mechanisms significantly improves latency, scalability, and security performance. This paper contributes a practical overview for both researchers and industry professionals aiming to design more resilient web architectures in the AI-driven digital era.

2. Machine Learning in Business Intelligence: From Data Mining to Strategic Insights in MIS

Authors: Shuvo, S. A.; Tabassum, M.; Tafannum, N.; Chadni, S.
Journal: Review of Applied Science and Technology, 4(2), 339–369 (2025).
DOI: https://doi.org/10.63125/drb8py41

Summary:
This journal paper focuses on how machine learning transforms the field of Management Information Systems (MIS) by enabling organizations to move beyond traditional data storage toward predictive and strategic decision-making. The research explains various data mining techniques and their applications in real-world business intelligence problems. It highlights supervised and unsupervised models used for market segmentation, sales prediction, customer profiling, and supply chain optimization. The paper emphasizes how adopting machine learning models can improve managerial insight and operational efficiency, ultimately leading to better strategic decisions. This study serves as a useful reference for professionals integrating AI and data science into business frameworks.

3. Effects of Citric Acid, Synbiotic, and Probiotic Supplementation Through Drinking Water on Growth Performance, Carcass Yield, and Blood Biochemistry of Broiler Chickens

Authors: Hossain, S.; Biswas, B. K.; Das, S.; Pory, F. S.; Raut, R.; Yeasmin, F.; Khan, S. U.; Dipta, P. M.; Shuvo, S. A.; Yeasmin, T.; Hoque, R.
Journal: Animals, 15(8), 1168 (2025).
DOI: https://doi.org/10.3390/ani15081168

Summary:
This interdisciplinary paper combines biological science and computational data analysis to evaluate how different natural supplements affect poultry growth and health. The study investigates the use of citric acid, synbiotics, and probiotics as eco-friendly alternatives to antibiotics in poultry production. By testing several groups under controlled feeding conditions, the research analyzes key indicators such as body weight gain, feed conversion ratio, carcass yield, and blood biochemical parameters. The results demonstrate that supplementing water with these natural additives improves nutrient absorption, boosts immunity, and enhances overall growth performance. The findings support sustainable agricultural practices and contribute to developing antibiotic-free livestock production models.

4. To Design & Develop Mobile-Based Birth Registration System for New Born Baby in Bangladesh

Authors: Shuvo, S. A.; Noyela, M. A.; Hossain, M. F.
Journal: International Journal of Software & Hardware Research in Engineering (IJSHRE), 9(6), 35–46 (2021).
DOI: https://doi.org/10.26821/IJSHRE.9.6.2021.9608

Summary:
This early work introduces a digital birth registration system designed to modernize and simplify the civil registration process in Bangladesh. The proposed web and mobile platform automates the process of registering newborns by linking parental identification numbers (NIDs) to a secure national database. The system generates a verified digital certificate, reducing manual paperwork and preventing data duplication. Developed using PHP, Laravel, and MySQL, the project represents a step forward in digital governance by making vital registration services accessible to rural populations. The study highlights how affordable and scalable digital tools can improve public services in developing countries.

5. Enhanced Smoke Detection Using Advanced Machine Learning Techniques: Integrating YOLO and KDE

Authors: Shuvo, S. A.; Liu, X.
Journal: IEEE Transactions on Image Processing (Under Review, 2025).

Summary:
This paper presents a novel approach to environmental smoke detection that combines object detection and statistical modeling. Using YOLOv7 for object localization and Kernel Density Estimation (KDE) for pixel-level refinement, the method achieves highly accurate and robust smoke detection in complex outdoor environments. The model was tested on drone and surveillance imagery of forest and urban regions. Results show a strong reduction in false positives compared to traditional detection systems. This research contributes to developing early wildfire monitoring tools that can prevent environmental disasters and protect natural ecosystems. It demonstrates how advanced computer vision techniques can be adapted for public safety and environmental protection.