Asian Journal of Research in Computer Science
Not a member yet
792 research outputs found
Sort by
Web Based Approach for SIWES Supervisors\u27 Reporting
No full textIn Nigerian universities, the Student Industrial Work Experience Scheme (SIWES) is a mandatory program designed to expose undergraduate students to real-world industrial practices relevant to their field of study. Despite the technological advancements in educational systems, the process of SIWES supervision and student performance assessment remains largely manual, inefficient, and error-prone. This project focuses on designing of a web-based system to enhance the supervision and grading process for SIWES students in the Department of Information Technology. The system enables supervisors to submit structured reports during field visits, assess students based on predefined criteria, and assign grades accordingly. Additionally, the platform facilitates the integration of departmental evaluations such as student defenses by combining defense scores with supervisor assessments to compute a final SIWES grade. The system was developed using the Laravel framework with HTML, CSS, JavaScript, and PHP. Through a thorough review of existing processes, system analysis, and software development methodologies, the project delivered a solution that improves accountability, streamlines communication, and enhances transparency in SIWES performance evaluation. The deployment of this solution is expected to improve the accuracy and efficiency of SIWES supervision and grading at both the field and departmental levels
Innovation Management in AI Development: Transforming Healthcare and Biopharma
No full textArtificial Intelligence (AI) is transforming healthcare and biopharmaceutical industries by revolutionizing diagnostics, personalizing medicine, and accelerating drug discovery. This study examines the critical role of innovation management in integrating AI technologies to drive value creation in these sectors. Through a comprehensive review of literature from 2017 to 2025, including peer-reviewed articles, industry reports, and case studies, we explore the applications, challenges, and opportunities of AI in healthcare and biopharma. The findings reveal that AI has the potential to significantly enhance diagnostic accuracy, streamline clinical trials, and reduce the time and cost of drug development. For instance, AI-powered tools like machine learning algorithms are improving disease detection through advanced imaging, while predictive analytics are enabling personalized treatment plans based on genetic and clinical data. In biopharma, AI is accelerating drug discovery by identifying potential drug candidates and optimizing clinical trial designs, as demonstrated by platforms like Atomwise and Insilico Medicine. However, the integration of AI into healthcare and biopharma is not without challenges. Ethical considerations, data privacy concerns, and the need for robust regulatory frameworks remain significant barriers. Issues such as algorithmic bias, the "black box" problem, and the lack of standardized data further complicate AI adoption. Effective innovation management is essential to address these challenges, ensuring that AI technologies are deployed ethically and efficiently. Strategies such as public-private partnerships, capacity building, and the development of open-source AI solutions are crucial for scaling AI in low- and middle-income countries (LMICs), where healthcare disparities are most pronounced. By addressing these challenges, AI can drive transformative advancements in patient care, therapeutic development, and global health equity, paving the way for a more efficient, personalized, and inclusive healthcare ecosystem
The Future of Cybersecurity: Predicting Trends and Preparing for Emerging Threats
No full textThe digital landscape of today’s world is evolving very rapidly, providing both prospects and hurdles, which calls for the vital importance of cybersecurity in the space of technological advancement. Focusing on future trends and emerging threats, this paper provides a forward-looking view to help organizations and individuals be prepared for the risks to come. As remote work, cloud computing, and IoT devices proliferate, traditional security solutions are ineffective. AI and machine learning innovations are changing how threat detection and prevention are being done; frameworks like Zero Trust are changing how safe access is being made. However, adversaries utilize these technologies to fit in highly sophisticated attacks, including AI-based malware and advanced social engineering techniques. We study the application of quantum computing in solving problems of existing cryptographic systems and the importance of post-quantum encryption protocols. The paper also talks about regulatory and ethical challenges and stresses the need for joint efforts between governments, organizations and researchers to design comprehensive security frameworks. This study identifies trends such as the move towards proactive threat intelligence and the combination of behavioral biometrics intended to deliver actionable insights for navigating the evolving cybersecurity landscape. The findings highlight continuous education, adaptive strategies, and investment into cutting-edge technologies to protect against tomorrow’s threat
WhatsApp Romanized Sinhala (Singlish) Group Chat Summarization Using NLP Techniques
No full textWith the growing popularity of WhatsApp group chats, especially in Sri Lanka, users increasingly face challenges of information overload, leading to missed or unread important messages. While solutions exist for summarizing English-typed messages, there has been no significant attempt to summarize Singlish, a unique typing style where Sinhala words are written using the English alphabet. This research aims to address this gap by developing a Natural Language Processing (NLP)-based system to automatically summarize Singlish-typed WhatsApp group chats over 24 hours. Using exported chat data without media attachments, a customized data pre-processing pipeline was developed to clean, tokenize, and extract keywords from the chats. Two popular Summarization models, facebook/bart-base and sshleifer/distilbart-cnn 12-6, were employed to generate concise summaries, which were then distributed to users via email. The system was evaluated through information retrieval metrics and human assessments to ensure relevance and quality. The study highlights the challenges of processing Singlish due to its informal variations and lack of language resources and sets a foundation for future improvements in chat summarization for low-resource languages. The developed solution not only enhances user productivity but also contributes to the broader field of localized NLP research
Stacked Boost Forest: A Hybrid Model to Predict Domestic Cinnamon Purchasing Cost in Down South Sri Lanka
No full textSri Lanka is the leading exporter of true cinnamon, providing 90% of global demand. However, domestic farmers face challenges in securing a stable market price due to varying prices set by different intermediate buyers and a lack of awareness of price fluctuation patterns. This research aims to develop a web-based forecasting system to predict the highest and average purchase prices of cinnamon from domestic farmers in southern Sri Lanka, using historical data from 2016 to 2024. The study introduces a hybrid model incorporating a Random Forest Regressor, a Gradient Boosting Regressor, and a Stacking Regressor with a Linear Regression meta-model, achieving 96% accuracy for the highest price prediction and 98% accuracy for average price prediction. Compared to previous studies that primarily focus on the export market, this research analyzes both external and internal factors influencing price fluctuations and considers both domestic and export markets. The proposed system provides stakeholders with a user-friendly platform to enhance price transparency and stability. Future work aims to expand the forecast coverage to the entire country and introduce a comparative report feature for year-over-year price analysis
Fine-Tuning DeepSpeech Speech-To-Text Model for Nigerian English and Yoruba-English Code-Switched Speech
No full textSpeech-to-Text (STT) systems, despite their stellar performance in recent years, still struggle with recognising non-Western English accents and speech that features Code-Switching (CS), a linguistic phenomenon common in regions such as Nigeria. This study addresses that challenge for Nigerian English and Yoruba-English code-switched speech by adapting Mozilla’s DeepSpeech 0.9.3 model and fine-tuning it using a custom dataset of 118 minutes (approximately 1.97 hours). This process involved transfer learning and hyperparameter optimisation over iterative training sessions on a CPU-based setup. The model’s performance was evaluated using Word Error Rate (WER) and Character Error Rate (CER), with the best model showing modest improvements over the baseline model and achieving a WER of 0.760261 and CER of 0.381241 after 55 epochs. Although limited computing resources and the small dataset imposed significant constraints on the work, the study demonstrated the potential of fine-tuning and transfer learning for model adaptation to low-resource languages and code-switching contexts. Future work will require access to GPU resources for improved convergence and transcription accuracy, an expanded dataset and support for Yoruba diacritics to improve the quality of transcriptions
Comparative Reliability Analysis of Selenium and Playwright: Evaluating Automated Software Testing Tools
No full textAims: This study aims to evaluate and compare the reliability of Selenium and Playwright, two leading frameworks for automated web testing. The assessment focuses on key reliability metrics, including uptime and the Rate of Occurrence of Failures (ROCOF).
Study Design: A comparative experimental study conducted under controlled testing conditions.
Place and Duration of Study: The study was conducted over a 24-hour continuous testing period using two laptops with distinct hardware configurations: an HDD-equipped HP laptop and an SSD-equipped Dell laptop.
Methodology: Reliability was measured using two core metrics: uptime and ROCOF. Tests were conducted on an HP laptop (HDD) and a Dell laptop (SSD). Two Python scripts — one for Selenium and one for Playwright — were developed to execute identical actions. For the 24-hour uptime test, Selenium ran on HP and Playwright on Dell. ROCOF was assessed at six-time intervals — 8:00 AM, 8:15 AM, 4:00 PM, 4:15 PM, 12:00 AM, and 12:15 AM — by alternating tool execution between HP and Dell, allowing for analysis of the effects of hardware and time of day on failure rates.
Results: Selenium achieved 100% uptime with no failures, while Playwright recorded 99.72% uptime with four downtimes. For ROCOF, both tools had one failure per 10-test sequence, but Selenium’s higher failure rate per second (0.1208 on HP, 0.1336 on Dell) was due to faster execution times (7.93s on HP, 7.87s on Dell) compared to Playwright (36.74s on HP, 35.84s on Dell). The SSD-equipped Dell laptop outperformed the HDD-based HP, with faster completion times (43.71s vs. 44.67s).
Conclusion: Selenium is ideal for scenarios requiring uninterrupted uptime, while Playwright\u27s consistent response times suit dynamic web application testing. The study highlights hardware\u27s role in performance, with SSDs offering superior speed and stability. These findings guide practitioners in choosing tools based on hardware, stability, and execution needs
Strengthening Compliance with Data Privacy Regulations in U.S. Healthcare Cybersecurity
No full textThis study evaluates the state of data privacy and cybersecurity compliance within the U.S. healthcare sector, leveraging data from the U.S. Department of Health & Human Services Breach Portal, Verizon Data Breach Investigations Report, and the Health IT Dashboard. A quantitative methodology comprising descriptive statistical analysis, K-means clustering, and multivariate regression was employed to examine healthcare data breaches, categorize cybersecurity threats, and identify compliance challenges. Findings revealed a persistent increase in breaches, with hacking/IT incidents comprising over 80% of breaches in 2020 and a peak of 135 incidents in 2021. Budget allocation emerged as the most significant predictor of compliance (p = 0.0178), affirming resource constraints. Malware and ransomware were identified as dominant threats, while insider threats emerged as high-impact vulnerabilities. The study recommends increasing cybersecurity budgets, implementing continuous staff training, harmonizing regulations, and adopting Cybersecurity Maturity Models to systematically enhance security postures. The study provides critical insights into the challenges faced by healthcare organizations in achieving compliance with evolving data privacy regulations such as HIPAA and HITECH. The findings highlight the economic and operational implications of non-compliance, including financial penalties, reputational harm, and patient trust erosion. The study further affirms the importance of strategic investments in advanced cybersecurity tools, policy harmonization, and employee education. Hence, policymakers and healthcare administrators can utilize these insights to foster a robust culture of compliance, ensuring the protection of sensitive patient information and the resilience of healthcare operations against cyber threats. The study suggests that future research explores integrating artificial intelligence, zero-trust architectures, and adaptive risk management frameworks to further enhance cybersecurity strategies and regulatory compliance
Pneumonia and COVID-19 Classification and Detection Based on Convolutional Neural Network: A Review
No full textPneumonia diseases are considered one of the most pandemic diseases by the WHO, claiming the lives of millions across the world. Therefore, the necessity of having mechanisms for early diagnosis and detection of these epidemic diseases to preserve people\u27s lives. On the other hand, the increase in cases requires not relying on traditional means of detecting diseases due to these tests\u27 limitations and high costs like RT-PCR and errors in interpretation by humans. Among the available methods for diagnosing Pneumonia diseases are X-rays and CT scans. For accurate and highly efficient diagnosis, computer-aided diagnosis (CAD) is required. Deep learning has been suggested as the best solution to enhance the prognosis for many lung diseases using CAD systems. The ability of machine learning algorithms was not sufficient to diagnose with high accuracy due to several challenges, like: limited dataset, high computational, and less robustness because of huge numbers of features in the images. The aspiration to deep learning to solve the problems of diagnosis more accurate, especially convolution neural networks, showed impressive results in the classification of images. The convolutions layer in the network with filters automatically discover the critical spatial and temporal features in an image. Hierarchical representations were used in human brains in designing the learning process of CNN. Several deep learning architectures have been used to detect pneumonia diseases, the most popular such as AlexNet, VGG-16, Inception-v1, Inception-v3, ResNet-50, Inception-ResNet-V2, and ResNet201. The strength of CNNs lies in the size of data to be trained. Previous algorithms have been used to pre-train on a vast dataset. An augmentation strategy has been used in the model design to increase the dataset\u27s size and quality artificially. This review aims to pave the way for better, more accurate, and efficient models for early detection and classification of lung diseases to improve patient survival
Detecting Diabetic Retinopathy Using Machine Learning Algorithms: A Review
No full textDiabetic retinopathy, a condition resulting from prolonged high blood sugar levels that damage the retina, can cause vision impairment and, if untreated, lead to blindness. With advances in medical imaging and the availability of fundus image collections such as Madrid Messidor and DRIVE, computer-aided diagnosis (CAD) systems have become instrumental in identifying and categorizing cases. Machine learning, a branch of artificial intelligence, has demonstrated remarkable success in medical image processing, showing great potential for the early detection of diabetic retinopathy—a condition often challenging to diagnose in its early stages due to a lack of symptoms. This review examines prior studies leveraging machine learning algorithms, such as convolutional neural networks (CNNs), support vector machines (SVMs), and k-nearest neighbors (KNN), for diabetic retinopathy detection using fundus image datasets. It also explores existing challenges, including dataset variability, computational demands, and the generalizability of models across diverse populations. Highlighting methodologies, datasets, and performance metrics like accuracy, sensitivity, and specificity, this article aims to provide a cohesive understanding of the current landscape, delineate strengths and limitations, and suggest directions for future research