Institutional-Repository, University of Moratuwa
Welcome to the University of Moratuwa Digital Repository, which houses postgraduate theses and dissertations, research articles presented at conferences by faculties and departments, university-published journal articles and research publications authored by academic staff. This online repository stores, preserves and distributes the University's scholarly work. This service allows University members to share their research with a larger audience.
 Research Publications |  Thesis & Dissertation |  E- Books |
 UoM Journal Publications |  UoM Conference Proceedings |  Articles published in Scimago's Q1 journals |  UoM Research Reports |  Other Articles authored by UoM staff |
Recent Submissions
item: Conference-Full-text
An E-learning system model to enhance user experience with content recommendation
(IEEE, 2024) Udugahapattuwa, DPD; Fernando, MSD
The rise of E-learning platforms has created a requirement to monitor and evaluate student performance while delivering interactive content, ultimately improving student learning. This research project focuses on studying the use of various data mining algorithms to extract user interactions from E-learning systems and identify patterns for recommending personalized content. The study will explore manipulating content through translations and formatting across different media to maintain high student interest. Additionally, it highlights the benefits of personalized learning, increased satisfaction, and early intervention in extracting student behavior. Moreover, it emphasizes best practices for formatting E-learning management system content, such as using headings, shorter paragraphs, images for illustration, and consistent style. Ultimately, this research concludes a model that helps to create an intelligent E-learning system that leverages data mining algorithms and machine learning techniques to generate personalized content recommendations based on user performance ratings to improve engagement and learning outcomes. In the initial testing of the model, it was given around 63.16% accuracy. After retraining the model, it was given a 78.90% accuracy in testing. Finally, the content will be arranged using the SCORM standard.
item: Conference-Full-text
A Novel lightGBM-bayesian approach for DDoS detection in SDN environments
(IEEE, 2024) Vaishali, R; Naik, SM
Software-Defined Networks (SDN) have revolutionized network management by introducing a centralized controller. However, this centralization renders SDN vulnerable to Distributed Denial of Service (DDoS) attacks, posing critical security challenges. While existing studies explore attack vulnerabilities in SDN, they often suffer from limitations related to memory management and efficient detection. To address these issues, we propose a novel model that leverages the Light-Gradient Boost Machine (LGBM) algorithm, coupled with Bayesian Optimization for hyperparameter tuning. Our model achieves an exceptional average accuracy of 99.18% on the UNSW-15 dataset during both training and testing phases, surpassing existing models in terms of accuracy and training time. By outperforming current solutions, our proposed DDoS attack detection model significantly enhances SDN security, providing a robust defense mechanism against emerging threats.
item: Conference-Full-text
Detection of tea leaf diseases using deep transfer learning
(IEEE, 2024) Vijayakanthan, G; Vaishali, R; Abolghasemi, V
Tea leaf diseases significantly impact both the quantity and quality of tea production in Sri Lanka, a country where tea cultivation holds considerable economic importance, contributing significantly to its GDP and serving as a major export to consumer markets. Existing computer vision and machine learning methods require a large number of image samples for accurate classification, leading to a time-consuming process. To address this limitation, we propose a novel approach utilizing deep transfer learning to train classification models efficiently with limited samples, leveraging cross-domain knowledge transfer. Our method aims to detect tea leaf diseases early, thereby preserving tea quality and fostering sustainable agricultural practices. The unique contributions of this study are a) collecting a comprehensive set of tea leaf images from different tea gardens representing six tea leaf conditions, annotated manually and b) developing a pre-trained convolutional neural network (CNN) architecture, with 256, 128, and 6 fully connected layers, including Xception, DenseNet201, VGG16, InceptionV3, EfficientNetB0, and MobileNetV2, to transfer classification knowledge. Through several experimentations with various fine-tuning techniques, we achieved a notable average accuracy of 99.58% in classifying tea leaf diseases.
item: Conference-Full-text
Redefining personal safety with innovative self-defense and emergency notifying technology through affective design
(IEEE, 2024) Galang, EM; Carungay, SM; Villamor, L; Cueva, MD; Cruz, RD
This study focuses on enhancing personal safety in the Philippines, specifically regarding street crimes like rape, assault, robbery, murder, pickpocketing, and theft. The objective is to develop a self-defense product that improves user safety through enhanced and innovative design and efficiency. Utilizing quantitative methods, the study examines factors influencing Filipinos' safety perception, identifies design improvements for existing self-defense tools, and assesses their impact on efficiency ratings. A comprehensive literature review underscores the importance of self-defense weapons and the need for innovative solutions within the Philippine context, and
the utilization of survey was used to consider the insights of the respondents on their environment, crime awareness, and experiences with self-defense tools. The findings inform the integration of effective features—such as stun guns, sound alarms, kubotans, flashlights, and an emergency notification tool—into the proposed product. A work measurement study involving participants experienced with conventional selfdefense keychains confirms the significant increase in efficiency ratings for the proposed product, validating the design improvements. The findings offer valuable insights into factors shaping safety perceptions and influencing the design of selfdefense weapons as the study highlights the proposed product as an innovative solution for improving the design and efficiency of self- defense weapons in the Philippine market, providing users with a reliable and versatile self-defense instrument and may paved way for further developing on integrating high- quality materials, refining location tracking capabilities, and incorporating intelligent features to comply with local regulations.
item: Conference-Full-text
Waste heat recovery from marine engines using absorption chillers for comfort application: a case study based on Hamilton and Saryu ship classes
(IEEE, 2024) Dikkumbura, R; Wijewardane, A
With the growing environmental concerns and the emission regulations already in place, as well as upcoming regulations in the future, there is a need to reduce emissions from marine vessels. The maritime industry is essential to the massive global movement of people and goods. Nonetheless, it is acknowledged that the maritime industry is responsible for 3% of the world's greenhouse gas (GHGs) emissions. The engine efficiencies of the modern marine ships are in the range of 30- 45%, highlighting that 70-55% of the fuel energy is discharged as waste heat to the surrounding atmosphere by the engine cooling system and the exhaust system. Energy efficiency
measures and stringent emission regulations have not been implemented in marine vessels and transport sector yet considering the cruise's reliability and safety due to the operating environment. However, the scope and the use of maritime transportation sector is expanding rapidly around the clock due to economic consideration and globalization. As mentioned above, approximately 2/3rd of the energy produced in the marine engines is wasted via the exhaust and the cooling system without being recovered. During the literature review, it was observed that theses mobile power plants used in maritime industry have not adopted waste heat recovery techniques adequately as much as the road transportation sector due to several reasons such that lack of attention to enhance the power train efficiency due to the lack of emission standards. Turbocharging, turbo-compounding, bottoming cycles (ORCs) and thermoelectric generation have been identified as the most promising and established energy recovery techniques that can be adopted to recover the waste heat and convert it to power in the marine vessels. However, no studies have been conducted to investigate waste heat recovery from marine engines using absorption chillers for air conditioning and passenger comfort in marine applications so far. Therefore, this study will investigate the applicability and possibility of exhaust energy recovery from absorption chillers to provide and handle the on-board cooling demand for thermal comfort applications. The study was conducted using the real engine data obtained from two classes of marine vessels (Hamilton class and Saryu class) use in Sri Lanka Navy. Results of the numerical analysis show that the heat which can be recovered from the main exhaust system of both classes of ships, is able to cater for the cooling load of the respective ship class. The analysis was performed based on LiBraq absorption refrigeration, which will ensure the selected air conditioning system does not exhaust GHGs during the operation.