By CSE Department https://ar.iub.edu.bd/handle/11348/643 2026-09-17T12:40:25Z 2026-09-17T12:40:25Z Hafiz, Zinia https://ar.iub.edu.bd/handle/11348/1568 2026-09-06T12:35:18Z 2025-12-14T00:00:00Z

A lightweight framework for implementing ISO/IEC 26550 in small and medium-sized software enterprises in Bangladesh Hafiz, Zinia This study explores the challenges faced by Bangladeshi software Small and Medium-sized Enterprises (SMEs) in adopting ISO/IEC 26550 for Software Process Improvement (SPI). Data collected from 12 SMEs revealed uneven implementation across key process areas, with high familiarity in domain requirements (up to 95%) but limited engagement in testing and validation practices. To address these gaps, the study proposes a lightweight PDCA (Plan–Do–Check–Act) framework tailored to the unique resource constraints of SMEs. Unlike traditional models such as CMMI or ISO/IEC 12207, the proposed approach emphasizes simplicity, cost-effectiveness, and role-based adaptability, making it more practical for SMEs in emerging economies. The framework supports scalable and gradual process improvement, enhances quality assurance, and facilitates stronger alignment with ISO/IEC 26550 in resource-limited environments. This thesis is submitted in partial fulfilment of the requirements for the degree of Master of Science (M. SC.) in Software Engineering, 2025.

2025-12-14T00:00:00Z Setu, Jahanggir Hossain https://ar.iub.edu.bd/handle/11348/1547 2026-08-24T08:35:25Z 2026-08-01T00:00:00Z

Declared failure over silent failure: abstention, explain ability and an abstention-aware evaluation framework Setu, Jahanggir Hossain Deep image classifiers are constructed in a way so that every input produces a label, whatever the model has learned about that input. The resulting failures are silent, e.g., a classifier confronted with an Out-Of-Distribution (OOD) or adversarially perturbed image returns a class assignment with no signal that the assignment should not be trusted. Selective classification yields a partial remedy by permitting the model to decline but its standard evaluation practice, e.g., computing performance over the accepted subset and reporting coverage separately, charges nothing for that decline. Therefore, a reported score improves monotonically as the model abstains more. This study makes two contributions. The first is empirical where a SelectiveNet architecture is compared against a matched baseline sharing the same convolutional backbone across four datasets (MNIST, CIFAR-10, chest X-ray for pneumonia detection and Cats vs. Dogs), evaluated, under Fast Gradient Sign Method (FGSM) perturbation and with Gradient-weighted Class Activation Mapping (Grad-CAM) applied to the selection head, as well as, the prediction head so that the abstention decision is itself explained. The second is the Abstention-Aware Evaluation Framework (AAEF) which extends the confusion matrix with an abstention column preserving every sample and defines three metrics on it. Boundedness, exact reduction to the classical metrics at zero abstention and correct extreme-case behavior are established and verified. Absolute error counts fell on all four datasets with missed pneumonia cases dropping from 17 to 7, though 4 to 14 deferrals were required per error avoided. Adversarial results were consistent, as well. As perturbation strength rises the baseline CNN degrades steadily, while SelectiveNet’s selective accuracy recovers because it lowers coverage and rejects the samples it cannot resolve. The optimism introduced by discarding abstentions ranged from 0.069 to 0.26 in balanced accuracy. Charging abstention as an outcome rather than an exclusion changes the reported performance of a selective classifier which is large enough to reverse a comparison that conventional evaluation would decide the other way. This thesis is submitted in partial fulfilment of the requirements for the degree of Master of Science in Computer Science, 2026

2026-08-01T00:00:00Z Khanam Mim, Morshada Jerin Arju, Mst. Tahmina https://ar.iub.edu.bd/handle/11348/1318 2026-07-04T11:05:15Z 2026-04-01T00:00:00Z

Investigating the factors affecting the intention to use AI Chatbots in STEM education app: a hybrid structural equation modelling and artificial neural network Khanam Mim, Morshada; Jerin Arju, Mst. Tahmina This study explores the impact of Artificial Intelligence (AI)-powered applications in Science, Technology, Engineering, and Mathematics (STEM) education, emphasizing their role in improving students’ problem-solving skills and self-efficacy. It examines how personal factors such as ICT self- efficacy and self-directed learning (SDL), along with technological aspects like perceived ease of use (PEU) and perceived convenience (PC), shape students’ engagement with AI-driven tools. Using a quantitative method, survey data were collected from 117 students during February–March 2025. The research employed the Structured Predictive Latent Semantic System (SPLSS) model and Artificial Neural Network (ANN), validated through Structural Equation Modeling (SEM), to ensure reliability and predictive accuracy. Results show that AI tools significantly enhance problem-solving and self-efficacy. PC and intention to use were strong predictors of chatbot utilization, while ICT self-efficacy and PEU influenced attitudes and behavioral intentions. Importance-Performance Map Analysis (IPMA) revealed convenience as most impactful, and self-efficacy as least. All hypotheses were supported, confirming the model’s robustness. The study concludes that AI-driven applications create personalized, engaging, and confidence-boosting STEM learning experiences, highlighting the need for user-friendly, contextually adaptive AI tools and suggesting future research on long-term impacts and scalability for inclusive STEM education. This thesis is submitted in partial fulfilment of the requirements for the degree of Master of Science (MSc) in Computer Science and Engineering (CSC), 2026.

2026-04-01T00:00:00Z Sayam, Muntakim Kadir https://ar.iub.edu.bd/handle/11348/1244 2026-08-20T10:35:14Z 2026-06-01T00:00:00Z

Bookplace: a modular & scalable multi-vendor ecommerce system with advanced order, inventory & analytics management Sayam, Muntakim Kadir On contemporary e-commerce platforms, the discovery of products, management of the cart, payment, and delivery are usually handled, whereas on professional home-service platforms, the focus is on technician booking, scheduling, and job completion. Customers who purchase appliances or products that consume services extensively typically require simultaneous workflows. A customer can buy an air conditioner, washing machine, smart gadget, or home appliance online, but must call a separate service provider, make a phone call, or use a third-party platform to book an installation, repair, or maintenance. This break creates conflicts with clients, ambiguity among suppliers, reduced visibility among technicians and issues with after-sales responsibility. This paper presents BookPlace, a service-commerce-based multi-vendor platform that offers a single modular web-based platform for online product and professional service booking. It includes product catalogs, technician schedules, receivables, loyalty services, reviews, refunds, support credentials, vendor processes, human resource operations, and administrative analysis to enable customers, vendors, technicians, and administrators to run the entire service-commerce life cycle more efficiently, safely, and collaboratively. This platform has been built with Laravel 12, Vue 3, InertiaJS, Tailwind CSS, MySQL, Spatie permissions, Laravel Fortify, Stripe, PayPal and SSLCommerz. It takes advantage of a modular monolith design, in which large business areas are subdivided into feature-oriented modules that remain part of a single deployable application. This thesis is submitted in partial fulfilment of the requirements for the degree of M.Sc. in Computer Science, 2026.

2026-06-01T00:00:00Z