Asian Journal of Research in Computer Science
Not a member yet
792 research outputs found
Sort by
Bridging Gaps in Cybersecurity Governance: Leveraging Collaborative Digital Solutions
No full textThis study investigates the effectiveness of collaborative digital solutions in addressing cybersecurity governance gaps within the educational sector, focusing on resource constraints, awareness deficits, and regulatory compliance challenges. Data from the National Institute of Standards and Technology Cybersecurity Framework Usage Dataset, the Global Threat Intelligence Sharing Alliance Data Repository, and the World Values Survey were utilized. Quantitative methodologies, including descriptive analysis, Difference-in-Differences, and logistic regression, were employed to analyze gaps, evaluate solution effectiveness, and explore barriers. Findings revealed significant governance gaps, with resource constraints showing a mean frequency of 140.3, the highest among categories. Collaborative solutions demonstrated a 49.3% reduction in breach incidents post-intervention. Logistic regression identified awareness as a major barrier with an odds ratio of 2.46. Recommendations include prioritizing cybersecurity awareness programs, enhancing access to collaborative solutions, standardizing data-sharing protocols, and investing in capacity-building initiatives to fortify institutional resilience
Personalization in Digital Marketing: Leveraging Machine Learning for E-Commerce
No full textIn the era of digital transformation, machine learning (ML) techniques have revolutionized personalized marketing, enabling businesses to enhance customer engagement through data-driven strategies. This paper presents a systematic review of ML applications in digital marketing and e-commerce, focusing on customer segmentation, recommendation systems, and targeted advertising. Specifically, it explores the role of collaborative filtering, deep learning, reinforcement learning, and hybrid AI models in improving personalization and predictive analytics.
The findings indicate that deep learning-based models, such as neural networks and transformers, significantly enhance personalization accuracy, while reinforcement learning optimizes real-time bidding and dynamic pricing strategies. Hybrid recommendation systems outperform traditional methods by combining user behavior data with contextual insights to improve ad targeting and customer retention.
Beyond theoretical insights, this study provides practical implications for marketers, data scientists, and e-commerce businesses, enabling them to optimize AI-driven personalization strategies for increased conversion rates and customer loyalty. However, challenges such as data privacy concerns, algorithmic biases, and high computational costs remain barriers to widespread adoption. Future research should focus on developing ethical AI frameworks to ensure fairness and transparency in automated personalization
Machine Learning Models for Predicting Parkinson’s Disease Progression Using Longitudinal Data: A Systematic Review
No full textThis systematic review aims to evaluate the effectiveness of various machine learning models in predicting PD progression using longitudinal data. Despite the increasing use of ML in PD research, gaps remain in understanding the impact of longitudinal data on prediction accuracy and model generalizability. This study aims to bridge this gap by examining how multimodal data sources, including clinical, genetic, and imaging datasets, contribute to improved predictive performance. The review focuses on the types of models used, data sources, performance metrics, and their potential to improve personalized treatment and clinical decision-making. A comprehensive literature search was conducted across Scopus, PubMed, Google Scholar, and ResearchGate to identify relevant studies published from January 2010 to February 2024. The inclusion criteria focused on studies employing ML techniques for analyzing longitudinal PD data, yielding 14 eligible studies. Data were extracted on ML models used, dataset characteristics, performance metrics, and the integration of multimodal data sources such as clinical, genetic, and imaging data. The findings were synthesized to assess model performance and generalizability. Long Short-Term Memory (LSTM) and ensemble methods like Random Forest and Light Gradient Boosting Machine (LGBM) are effective in capturing disease progression with high accuracy and robust performance metrics. LSTM models achieved accuracies up to 90% and AUC scores of 93.79%, while LGBM models achieved 90.73% and AUC of 94.57%. The Matthews Correlation Coefficient (MCC) scores in longitudinal studies increased over time, and Mean Absolute Error (MAE) also improved. Integrating multimodal data, including clinical, genetic, and imaging information, further improved model reliability and generalizability. ML models, particularly those incorporating longitudinal and multimodal data, show promise in predicting PD progression. Future research should prioritize dataset diversity, enhance model interpretability, and leverage real-world wearable data for improved clinical applicability
Customer Churn Prediction in the Telecommunication Industry Over the Last Decade: A Systematic Review
No full textAims: This study explores the application of machine learning algorithms in predicting customer churn within the telecommunications sector. By analyzing various predictive models, the study identifies key factors influencing churn and assesses how data integration enhances predictive accuracy.
Study Design: A systematic literature review was conducted to evaluate existing research on churn prediction models and their effectiveness in the telecommunications industry.
Place and Duration of Study: The study reviews published research from various academic and industry sources over the past decade, focusing on global trends in customer churn prediction.
Methodology: Relevant studies were systematically selected and analyzed based on predefined inclusion criteria. The review examined different machine learning techniques, predictive variables, and data sources used to improve churn prediction accuracy. Special attention was given to real-time data integration and the impact of external datasets on model performance.
Results: Findings indicate that data integration, particularly real-time and external data sources, significantly enhances churn prediction accuracy. Machine learning techniques, including traditional models and emerging deep learning approaches, show promising results in improving customer retention strategies. However, challenges such as data privacy concerns and the need for methodological advancements remain. The study recommends further exploration of deep learning models to refine predictive capabilities and support robust retention strategies in the telecommunications sector
Enhancing Site Reliability Engineering Through AIOps: A Framework for Next-Generation IT Operations
No full textThe increasing complexity of modern IT infrastructures has pushed traditional operational approaches beyond their limits. This paper explores the integration of Artificial Intelligence for IT Operations (AIOps) within Site Reliability Engineering (SRE) practices to address this challenge. I present a framework for enhancing core SRE concepts such as Service Level Objectives (SLOs), Service Level Indicators (SLIs), and error budgets through AI-driven capabilities. Our approach enables more dynamic reliability targets, intelligent anomaly detection, and automated remediation while maintaining the engineering rigor of SRE. Case studies demonstrate significant improvements in key operational metrics: 87% reduction in alert noise, 73% decrease in mean time to detection, and 62% of common infrastructure issues resolved automatically. The proposed framework provides a systematic path for organizations to evolve from traditional SRE to AI-enhanced reliability practices while addressing common implementation challenges including data quality issues, skills gaps, and organizational resistance. This integration represents a fundamental shift in IT operations from reactive human-centered approaches to proactive AI-augmented engineering disciplines capable of managing unprecedented scale and complexity.
Aims: To develop and validate a framework that integrates Artificial Intelligence for IT Operations (AIOps) within established Site Reliability Engineering (SRE) practices, addressing the growing complexity of modern IT infrastructures.
Study Design: A mixed-method research approach combining case studies, controlled experiments, and quantitative analysis across multiple industry sectors.
Place and Duration of Study: The research was conducted across three major organizations in financial services, healthcare technology, and e-commerce sectors between January 2023 and February 2024.
Methodology: I developed an integrated framework enhancing five core SRE functions with AI capabilities. Implementation followed a four-phase methodology addressing technical, process, and organizational aspects. Effectiveness was measured through comparative analysis of key operational metrics pre- and post-implementation, including alert volumes, detection times, resolution rates, and operational burden.
Results: Implementation demonstrated significant operational improvements across all organizations. Key results include: 87% reduction in alert noise while maintaining critical issue coverage, 73% decrease in mean time to detection for system anomalies, 62% of common infrastructure issues resolved automatically without human intervention, and 47% reduction in SRE on-call burden. The financial services organization identified five previously unmonitored SLIs that significantly impacted user experience, while the e-commerce platform successfully predicted capacity-related incidents 30-45 minutes before impact.
Conclusion: The integration of AIOps with SRE practices creates a powerful combination capable of managing the scale and complexity of modern IT environments. The framework enables organizations to progress from reactive to predictive operations while maintaining the engineering rigor of traditional SRE. Future research should explore incorporating emerging technologies such as large language models and developing industry-specific implementations for sectors with unique reliability requirements
The Transformation of Business Processes: Harnessing the Power of Artificial Intelligence, Machine Learning, and Blockchain
No full textThe way businesses operate is changing at an unprecedented pace, thanks to the rapid evolution of digital technologies. Among the most transformative are Artificial Intelligence (AI), Machine Learning (ML), and Blockchain—three powerful tools that are revolutionizing industries by making processes smarter, faster, and more secure. AI and ML have shifted the way companies make decisions, allowing them to harness data for automation, predictive insights, and enhanced customer experiences. From chatbots that provide instant support to algorithms that optimize supply chains, these technologies streamline operations and drive efficiency. Meanwhile, Blockchain is redefining trust in digital transactions, ensuring data integrity, transparency, and security—especially in finance, supply chain management, and identity verification. When combined, these technologies do more than just improve processes; they create a foundation for businesses to scale, cut costs, and adapt to an increasingly digital world. As organizations continue to integrate AI, ML, and Blockchain, they face both opportunities and challenges. This paper explores how these innovations are shaping modern business processes, their real-world applications, and what the future may hold
Explainable AI in Regulatory Compliance: Balancing Transparency and Performance in AI Driven Treasury Management
No full textThe integration of Artificial Intelligence (AI) in financial operations has transformed treasury management by enhancing efficiency, risk assessment, and compliance processes. However, the increasing use of AI in regulatory compliance introduces a critical challenge: balancing transparency and performance. Explainable AI (XAI) has emerged as a solution to enhance interpretability, accountability, and trust in AI-driven decision-making. This article explores the role of XAI in regulatory compliance for treasury operations, addressing key challenges and trade-offs between explainability and performance. It evaluates existing frameworks, regulatory expectations, and industry best practices for implementing XAI in financial institutions. Additionally, this study highlights the impact of explainability on AI model efficiency and proposes strategies to optimize performance without compromising compliance. The findings provide valuable insights for financial professionals, regulators, and AI developers seeking to navigate the evolving landscape of AI-driven treasury management
A Hybrid PhoBERT-CNN-LSTM Model for Sentiment Analysis of Vietnamese Student Feedback
No full textStudent feedback plays a crucial role in improving educational quality and creating an effective learning environment. This study applies sentiment analysis to Vietnamese student feedback to extract and classify emotions into positive, negative, and neutral categories, providing valuable insights to support teaching improvements. We propose a method that utilizes the PhoBERT model for semantic feature extraction, followed by a CNN-LSTM architecture to capture both local features and sequential relationships in feedback data. Experimental results on the UIT-VSFC dataset demonstrate that the proposed PhoBERT-based CNN-LSTM model achieves an accuracy of 93.24% and an F1-score of 92.92%. This model demonstrates superior performance compared to several other advanced approaches. It surpasses ensemble model, which achieved an F1-score of 92.79%. These findings confirm the effectiveness of the model in extracting and classifying sentiments from student feedback while proposing a practical approach for analyzing Vietnamese educational data, contributing to teaching quality enhancement
A Hybridized Machine Learning Based Crisis Period Prediction System for Epileptic Patient Using Crisp-DM and SVM
No full textEpileptic seizures are unpredictable and can severely impact the quality of life of patients. To address this challenge, this research presents a hybridized machine learning-based crisis period prediction system designed to predict seizure occurrences with high accuracy. The system leverages a comprehensive dataset that integrates physiological signals, environmental variables, and behavioural patterns, offering a holistic approach to seizure prediction. The methodology follows the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework, ensuring a structured approach to data pre-processing, model development, and evaluation. Object-Oriented Analysis and Design Methodology (OOADM) principles were employed to modularize and optimize the implementation, facilitating scalability and maintainability of the system. The dataset underwent rigorous pre-processing, including normalization, feature selection, and handling of missing values, to ensure the quality and reliability of the input data. A Support Vector Machine (SVM) classifier was employed due to its robustness in handling high-dimensional data. The evaluation of the model yielded an accuracy of 85%, demonstrating its effectiveness in predicting crisis periods for epileptic patients. This system represents a significant step forward in predictive healthcare for epilepsy management. By integrating diverse data sources and leveraging advanced machine learning techniques, it offers a promising tool for real-time crisis period prediction, potentially improving patient safety and autonomy. Future work will focus on enhancing the model\u27s accuracy and integrating real-time monitoring capabilities to enable proactive interventions
A Comprehensive Analysis to Detect Chronic Kidney Disease and Stage Prediction: Using Machine Learning
No full textChronic Kidney Disease (CKD) is worldwide health concern that causes other diseases and has a high rate of morbidity and mortality. CKD often progresses without noticeable symptoms in its early stages. Early detection is critical to delay or prevent progression to kidney failure. Traditional detection relies on lab tests (e.g., serum creatinine, GFR, urinalysis). Machine learning and Deep learning model can complement these methods by identifying subtle patterns in clinical or imaging data. This article investigates the applications of ensemble learning methods, such as AdaBoost, Random Forest, Gradient Boosting, and Voting Classifiers, for CKD prediction. These models address significant problems with CKD datasets, including missing values, unbalanced classes, and excessive complexity, by utilizing the benefits of several base learners. The ensemble learning models provide a tool for the early detection of CKD and tailored treatment by effectively representing the different patterns and interactions observed in the data