Browsing by Author "Wijendra, D. R"

Now showing 1 - 6 of 6

Embargo
Application of the Refactoring to the Understandability and the Cognitive Complexity of a Software
(IEEE, 2022-07-18) Wijendra, D. R; Hewagamage, K. P
Cognitive complexity of a software determines the methodology of comprehending the internal logic of a given software by an individual, quantitatively. The procedure of handling a software by different users is different, which results the cognitive complexity as a subjective measurement. The quantification of the cognitive complexity is still not standardized due to the varied number of factors affected for the cognitive complexity determination and its nature of the subjectivity. This paper evaluates the relationship between the cognitive complexity and understandability as one of the qualitative factors to determine the cognitive complexity and the usage of refactoring techniques to reduce the cognitive complexity without refraining its calculation process with respective to the internal logic of the software as in other standard software complexity metrics perform.
Embargo
Cognitive Complexity Reduction through Control Flow Graph Generation
(IEEE, 2022-07-18) Wijendra, D. R; Hewagamage, K. P.
The cognitive complexity of a software determines the human comprehension effort to determine its underlying logic. The human comprehension effort preliminary depends on the person, who deals with the software, the source code and the problem to be addressed. The understandability of a given source code is varying with each user such that the cognitive complexity results with a subjective measurement. The graphical representation of the logical behavior of a source code implies the individual to comprehend the logic easily rather than referring to its original code base. This paper evaluates the possibility of using control flow graph representation to reduce the cognitive complexity of a given source code, thereby ensuring a less software complexity.
Embargo
Digital Democracy: A Secure Platform for Voting
(IEEE, 2022-01-17) Peiris, K.T. I. U; Gunathilake, G. R. T; Attanayaka, J. A. E. P; Ilankoon, I. M. D. D; Chandrasiri, S. S; Wijendra, D. R
The voting system is the way to elect country representatives. It is a set of rules that determine how the elections and referendums are conducted and how their results are determined. Voting is commonly related to politics and is finished with exploitation and a manual approach. The voters stand to vote for his or her decision. There are many ways to conduct voting, and for a developing country like Sri Lanka, the polling power to have an excellent representative to develop the country into a better and prosperous one. Usually, the existing voting system is operated manually and sometimes may lead to malpractices. Many have suggested diverse ways to overcome the problems, and sometimes those recommendations have failed. The most reliable way to accomplish this is to expand the technology from manual voting systems to digital voting systems. There are several significant issues with introducing such a digital voting system. A significant problem here is the low level of technical knowledge among the people of Sri Lanka. Such a system would be a problem for some at once. This research paper aims to put forward a study on why and how it is necessary to shift from a standard voting system to a digital platform and how it saves time and energy.
Open Access
Optical Insight: Enhancing Ophthalmic Diagnostics with Automated Detection of Retinal Abnormalities
(International Association of Computer Science and Information Technology, 2025-06-11) De Silva, D.I; Wijendra, D. R; Siriwardana, K.S; Gunasekara, S.N.W; Piyumantha, U; Thilakaratne, S.P
Early and accurate detection of retinal diseases is crucial for preventing vision loss, yet traditional diagnostic methods remain limited by subjectivity and inefficiencies. This study introduces an Artificial Intelligence (AI)-driven diagnostic system leveraging hybrid deep learning models to detect Glaucoma, Macular Hole, Central Serous Retinopathy, and Drusen using fundus images. By integrating multiple architectures, including Residual Network (ResNet), Visual Geometry Group 16-layer network (VGG16), Densely Connected Convolutional Network (DenseNet), U-shaped Network (U-Net), and You Only Look Once version 8 (extra-large variant) (YOLOv8x), the system enhances diagnostic precision and generalization across diverse imaging conditions. Key innovations include the hybrid ResNet-VGG16 and DenseNet-VGG16 models, which significantly improve detection accuracy for Drusen and Central Serous Retinopathy, respectively. Additionally, the U-Net-ResNet hybrid architecture mitigates overfitting, ensuring more reliable Macular Hole detection, while the YOLOv8x object detection model outperforms traditional approaches in Glaucoma localization by accurately identifying the optic disc. These models, integrated into a web-based diagnostic platform, achieved sensitivities and specificities exceeding 95%, establishing a new performance benchmark for automated ophthalmic diagnostics. This research advances medical image analysis by demonstrating the efficacy of hybrid deep learning models, offering a scalable AI solution for early retinal disease detection. Its integration into clinical workflows highlights its potential to transform ophthalmic care, enhancing accessibility and improving patient outcomes.
Embargo
PROBEXPERT: An Enhanced Q&A Platform for Reducing Time Spent on Learning and Finding Answers
(IEEE, 2022-07-18) Thennakoon, K; Ekanayake, D; Marapana, T; Ranasinghe, A; Wijendra, D. R; Gamage, A
The World Wide Web contains a wide range of material from a variety of fields. However, when concerns towards the computer science domain, information users find on the internet may not be up-to-date due to the rapid pace of change and having to spend less time on the internet for researching and debugging tasks is an added luxury. Having an expertise level while providing answers through a platform is convenient for users, yet when a user signs into a platform, the user must start from the beginning, regardless of the level of competence in the field. Moreover, not having a proper way to evaluate the existing programming knowledge is another obstacle. To address mentioned complications, researchers of this paper have introduced a new e-learning platform- ‘ProbExpert’. The platform has been constructed with machine learning and deep learning approaches such as NLP, keyword extraction, semantic information analysis, cosine similarity, and information summarization. With aforesaid technologies, ProbExpert provides systems in automated answering, optimized answer generation, structured question-based quiz evaluation together with a fully automated portfolio generation with a novel user ranking algorithm based on the bell curve.
Embargo
Virtual Dressing Room: Smart Approach to Select and Buy Clothes
(IEEE, 2021-12-09) Weerasinghe, S. W. P. N. M; Rajapaksha, R. M. D. D; Sathsara, L. G. I; Gunasekara, H. S. D. N; Wijendra, D. R; De Silva, D. I
The clothing industry portrays a major part of a respective country`s economy. Due to the predilection for clothing items of the people have led to the increasing of physical and online clothing stores in all around the world. Most of the people are used to go to the physical shopping and purchase their desired clothing items. But, as a consequence of the current pandemic situation, most of the people are unable to step out from their homes. This application is intended to cater an opportunity to the customers, who are not able to reach the physical clothing stores due to a pandemic situation and mobility difficulties. In addition, this application diminishes the time wastage, clothing size mismatches and the lesser user satisfaction ratio inside a physical clothing store. A customized 3D model has featured in the application to cater the virtual fitting experience to the customer. And the AI chatbot assistant in the application interacts with the user while catering virtual assistance for a better cloth selection process. In addition to that, this application has concentrated on the clothing shop by providing a future sales prediction component utilizing the K-Nearest Neighbors algorithm to provide an aid to their business commitments.