Theses

Now showing 1 - 3 of 3

Open Access
Design and Simulation of a Secure Enterprise IoT Network Using Cisco Packet Tracer with a Federated Learning-Based Secure Method.
(Sri Lanka Institute of Information Technology, 2025-12) Sumanadeera, H.M. G. G D
The rapid growth of the Internet of Things (IoT) in businesses has led to major security issues, with botnet attacks being a serious threat. Although Federated Learning (FL) provides a way to detect threats while preserving privacy, it is still vulnerable to data poisoning from harmful devices. Current blockchain solutions for securing FL are often too heavy on resources for widespread use in IoT. This paper offers a two-part integrated approach. First, an enterprise IoT network was designed and simulated securely using a prototype with Cisco Packet Tracer. Second, a new lightweight novel FL framework was developed that did not rely on blockchain, using the N-BaIoT dataset to protect against botnet attacks. The paper proposed Reputation- Weighted Coordinate Median with Update Validity Tests (RWCM+UVT) framework incorporates a reputation-based system, a robust aggregation algorithm, and an adaptive update validation gate. By simulating botnet attacks within this controlled environment, this paper demonstrates that the RWCM+UVT framework effectively identifies and mitigates the impact of malicious devices, achieving near-perfect detection accuracy without the prohibitive overhead of blockchain technology.
Open Access
Design and Implementation of an AI-Assisted Code Review Tool for Low-Code Platforms to Improve Quality and Security
(Sri Lanka Institute of Information Technology, 2025-12) PATHIRANA P.P.P.S.P
Low-code platforms like Mendix fast-tracks application development but, due to limited review mechanisms, face challenges in sustaining the code quality and security. Existing code review approaches are not optimized for visual cues, model-driven workflows, increasing the possibility of logical, security, and performance issues introduced by citizen developers. This research introduces an AI-assisted code review tool that combines GPT-4 and Claude Opus 4 for workflow analysis and defect detection in low-code environments. The approach evolved from few-shot prompting to workflow-oriented fine-tuning, resulting in improved analytical precision and reliability. The tool was further enhanced to perform business gap assessments and deliver user-friendly, structured feedback via a pluggable React-based widget integrated into the Mendix environment. The evaluation of the tool demonstrated an average precision of 84.5% and an average recall of 84.8% and an F1 score between (0.82-0.87), with workflow-based fine-tuning outperforming few-shot learning. A preliminary usability study with 25 developers demonstrated a 90% satisfaction rate and approximately 50% reduction in issue resolution time. Proxy validation using generative AI models was performed due to the limited availability of Mendix domain experts. These findings highlight the capability of AI-assisted code review to enhance workflow quality, strengthen application security, and improve developer productivity in low-code environments.
Open Access
Design and Validation of an AI-Enhanced Career Guidance Framework for Sri Lankan Secondary Education
(Sri Lanka Institute of Information Technology, 2025-12) Karthiha, S.
Career guidance is a decisive factor that contributes to the educational and professional path of students, but in Sri Lanka, the practices that are currently being used are mostly manual, fragmented, and unfair. This paper forms a conceptual model of an Artificial Intelligence (AI)-driven career guidance system that provides students with personalized educational and professional advice to high school students. The study is based on a mixed-methods research framework that combines both quantitative data collected in the form of the survey of 379 educational professionals working in nine provinces and the qualitative information gathered in the form of the interviews with ten experts. Key determinants of career decision-making were identified in a quantitative analysis and found to include academic performance, aptitude, personal interests, and socio-economic background, and six interrelated dimensions were identified in a qualitative thematic analysis: human-dominated guidance, family and cultural influence, perceived fairness of AI systems, ethical and infrastructural barriers, hybrid human-AI collaboration, and equity of access. The presented framework integrates these insights into a hybrid structure, which will combine AI-based analytics with advisory and parental judgment and will be culturally sensitive and ethically valid. The model is more accurate, inclusive, and efficient because it matches student profiles with the trends of the labour-market. The results prove that AI is capable of being more of a supplement to the human knowledge and can be used to increase access to data-driven counselling in even schools with scarce resources. The research has theoretical value in that it links AI technology to the career development theory and practical in that it will provide a replicable solution to the modernization of school guidance ecosystem in Sri Lanka by policymakers and educators. It ends with implementation recommendations on data ethics, transparency and capacity building to attain equitable and evidence based educational and career decisions support.

Theses

Browse

Filters

Advanced Search

Filter by

Settings

Sort By

Results per page

Search Results