		FeedBack \| Contact Us \| Links \| Site Map

Welcome to Journal of Theoretical and Applied Information Technology

Submit Paper / Call for Papers

Journal receives papers in continuous flow and we will consider articles from a wide range of Information Technology disciplines encompassing the most basic research to the most innovative technologies. Please submit your papers electronically to our submission system at http://jatit.org/submit_paper.php in an MSWord, Pdf or compatible format so that they may be evaluated for publication in the upcoming issue. This journal uses a blinded review process; please remember to include all your personal identifiable information in the manuscript before submitting it for review, we will edit the necessary information at our side. Submissions to JATIT should be full research / review papers (properly indicated below main title).

Journal of Theoretical and Applied Information Technology
April 2025 | Vol. 103 No.7

Title:	HEART FAILURE DETECTION USING OPTIMIZATION ALGORITHMS
Author:	AMARAVARAPU PRAMOD KUMAR , YUVARAJ MACHA , A SIVA KUMAR , BUNA SEKHAR , SINGAMANENI KRISHNAPRIYA , and Dr.S CHANTI
Abstract:	Heart failure (HF) remains a significant global health challenge, requiring early and precise detection to improve clinical outcomes and reduce mortality rates. Traditional diagnostic approaches often fail to capture the complexity of HF pathophysiology, necessitating advanced computational methods for accurate prediction. In this study, we propose a novel optimized Stacked Support Vector Machine (S-SVM) frame- work, integrating multiple SVM classifiers with diverse kernel functions to enhance predictive accuracy. A genetic algorithm (GA) is employed to fine-tune hyperparameters, ensuring model robustness and general- izability across patient populations. The model is rigorously evaluated on the UCI Heart Failure Clinical Records Dataset and the Framingham Heart Study Dataset, demonstrating superior performance in accuracy (95.7%), precision (0.90), recall (0.87), and AUC (0.96) compared to conventional machine learning tech- niques. The proposed system effectively balances computational efficiency with clinical interpretability, making it a promising tool for early-stage HF detection and risk stratification. This research advances the intersection of machine learning and cardiovascular diagnostics, offering a scalable and adaptive solution for real-world healthcare applications.
Keywords:	Heart Failure Prediction, Stacked SVM, Genetic Algorithm, Machine Learning, Clinical Deci- Sion Support, Cardiovascular Diagnostics.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	A HYBRID DEEP LEARNING FRAMEWORK FOR ACCURATE AND EFFICIENT DETECTION OF POTATO LEAF DISEASES
Author:	GIRIGULA DURGA BHAVANI , MUKKOTI MARUTHI VENKATA CHALAPATHI
Abstract:	Despite significant advancements in deep learning-based plant disease detection, existing models often struggle with generalizability, computational efficiency, and adaptability to real-world agricultural challenges. To address these limitations, this study introduces a novel hybrid deep learning framework combining the custom DarkPotatoleafNet model with Vision Transformers (ViT) and convolutional neural networks (CNNs). Our approach leverages transfer learning and ensemble learning techniques to enhance disease classification accuracy while optimizing computational resources. The highest recorded accuracy of 99% was achieved through this hybrid architecture, demonstrating its superior performance over conventional models. To ensure optimal model training and generalization, we employed extensive preprocessing techniques, including data augmentation (rotation, scaling, flipping), grayscale conversion for structural emphasis, and image masking and segmentation to isolate diseased regions accurately. This research contributes to precision agriculture by providing a scalable and efficient AI-driven solution for early and accurate detection of potato leaf diseases, ultimately supporting sustainable farming practices and improved crop yield management.
Keywords:	Transfer Learning, Sustainable Farming, Image Processing, Hybrid Models, Comparative Analysis.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	DYNAMICS OF DEPLOYING EDGE COMPUTING IN SOUTH AFRICAN HIGHER EDUCATION INSTITUTIONS
Author:	PHILSIWE SITHOLE , TENDANI LAVHENGWA
Abstract:	The purpose of this study was to conceptualize an edge computing deployment framework for the South African Universities context. The study argues that South African Higher Education Institutions (HEI) systems can improve by introducing technologies such as fog and edge computing to connect the smart devices that are shaping the present and future. This was to improve efficiency in day-to-day HEI operations and to gain a competitive advantage. To assist in understanding the adoption of edge computing, Unified Theory of Acceptance and Use of Technology (UTAUT), Task-Technology-Fit (TTF) and Media Research Theory (MRT) theories were used as research lenses. Other relevant tools and technologies for deployment, such as Wi-Fi, smart devices, and big data storage, are derived from data collected. The model is envisaged to improve the adoption, deployment and use of edge computing in university settings.
Keywords:	Edge Computing, Higher Education Institutions, UTAUT, TTF and MRT.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	SHRIMP CLASSIFICATION VIA MULTI-VIEW FEATURE EXTRACTION USING PRETRAINED DEEP FEATURES
Author:	P.V. NAGA SRINIVAS , Dr. M.V.P. CHANDRA SEKHARA RAO
Abstract:	Shrimp classification is a complex task in computer vision applications. Very few research works have been carried out by researchers because shrimp features are difficult to extract due to their complex shape. The taste of shrimp-based foods depends on whether the same shrimp category is used. The export shrimp business relies on the quality of classification. Manual identification and classification of shrimps is a time-consuming process. In this work, Shrimps are classified based on different angular views, shape, and texture features. Features are extracted using both Zernike and pre-trained deep features. Further, shrimps are classified by regression. The qualitative and quantitative analysis shows that the proposed framework performs better with hand-crafted and deep features. The proposed framework with Zernike moments and VGG-16 features shows robustness with an accuracy of 89% and a precision of 91.7%.
Keywords:	Shrimp classification, multi-feature extraction, Zernike features, VGG-16, multi-view features.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	STUDENT ACADEMIC PERFORMANCE PREDICTION USING PROBIT-GENERATIVE RESTRICTED DEEP BOLTZMANN MACHINE
Author:	C. VIVEK , P.TAMILSELVAN
Abstract:	Big data mining also known as Knowledge Discovery in Databases (KDD) is a process that uses statistical and computational techniques to identify significant hidden patterns and insights in massive datasets. The prediction of students' academic achievement is a key component of modern educational systems that aim to raise educational standards. Even though predicting students' performance has been the focus of numerous studies, there are still problems with early dropout prediction, learning pattern recognition, and student success prediction, which cause delays and reduced efficiency. To overcome these issues, Probit Generative Restricted Deep Boltzmann Machine (PGRDBM) method utilizes Deep Boltzmann Machine to include four different layers such as one input layer, two hidden layers, and one output layer. At first, the data collection followed by efficient performance of two significant stages such as feature selection and data classification to enable more accurate and time, space-efficient student’s academic performance prediction in educational institutions. At the input layer, the process of analysing educational performance begins with the collected student data. The Concept Drift-based Feature Selection (CD-FS) model is designed at the hidden layer to choose most pertinent features that are stable and relevant even in drift for precise prediction. The Probit Regression-based Data Classification (PR-DC) model is applied for robust student academic performance prediction by student data and classifying student performance into different classes. Then, hidden layer 2 sends the prediction results to the output layer. It is feasible to predict academic performance effectively. The performance analysis is effectively proposed in student academic performance prediction and two existing methods are conducted on different specific metrics such as accuracy, time, error rate, and space complexity
Keywords:	Big data Mining, Educational Performance Analysis, Prediction Achievement, Concept Drift, Feature Selection, Probit Regression and Data Classification.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	DESIGNING AND IMPLEMENTING A MULTI-CRITERIA INTELLIGENT FRAMEWORK TO ENSURE THE RELEVANCE OF REQUIREMENTS SOURCED FROM SOCIAL NETWORK SITES
Author:	MEKURIA SINKIE, TOR MORTEN GRONLI, DIDA MIDEKSO, ABDULLAH LAKHAN
Abstract:	The proliferation of social networking sites (SNS) and software failures primarily arising from the requirements elicitation phase, motivated researchers to develop methods, that incorporate SNS-based users’ needs into the requirements engineering stage, crucial for developing reliable software. This approach improves user-centric needs and identifies innovative features, but relevance verification has not been thoroughly examined, leading to challenges in filtering and prioritizing relevant information emanated from jargon, informal language, and diverse expressions detected in user comments. This research proposes a novel intelligent framework for relevance verification of SNS-sourced requirements, combining multiple criteria like organizational goals, business rules, related service datasets, and user comments. The proposed framework balances user, organization, and developer needs by simplifying the process of determining relevant requirements and enhancing their validity. The framework learns and utilizes consolidated criteria features as regulation mechanisms, addressing challenges in isolating relevant users' needs and minimizing traditional method limitations. The study uses qualitative methods for framework development and empirical research methods for sentiment and trend analysis, combining customized word embedding models and natural language processing. A case study of digital healthcare systems in developing nations explores three evaluation categories: business rules, related service datasets, and a blend of both. The dataset includes 2400 key phrases from 800 user needs, 540 business goals, and 900 from service datasets. The proposed method achieved a relevance rate of 88%, surpassing individual methods. The study contributes to IT and software engineering fields by providing a novel framework for relevance verification of SNS-based requirements, ensuring their alignment with actual user needs and improving their completeness and prioritization, leading to significant enhancements in system design and development.
Keywords:	Requirements relevance, Users’ need, Intelligent model, Word embedding and NLP, SNS
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	STRATEGIC OBJECTIVES FOR INNOVATIVE DEVELOPMENT OF MARKETING AND E-COMMERCE AS PART OF DIGITALIZATION OF CORPORATE BUSINESS
Author:	ILONA BABUKH , NARINE AVANESYAN , SVITLANA POVNA , OLEKSANDR BILANENKO, VIACHESLAV SAKUN, DMYTRO ABAKUMOV
Abstract:	The article is devoted to study of strategic objectives of innovative development of marketing and e-commerce as part of global digitalization of the enterprises’ business. Pros and cons of digital marketing and e-commerce providing for optimized and increased business efficiency are analyzed. Key barriers to development of e-commerce market and digital marketing are outlined. It has been proven that to increase the competitiveness level and to accelerate innovative development of business, enterprises should actively use e-commerce tools combined with digital marketing. Global trends in development of digital marketing and e-commerce were analyzed, as follows: indicators of the digital marketing use and forecast for the near future, dynamics of e-sales, sales of retail e - commerce worldwide, channels used in marketing, enterprises using e - business applications. Strategic priorities of innovative development of marketing and e-commerce as part of business digitalization of enterprises are substantiated.
Keywords:	Strategy, Strategic Development, Innovative Development, Innovative Potential, Marketing, E-Commerce, Digitalization, Digital Technologies, Competitiveness, Investments, Corporate Business, Enterprise.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	ZERO TRUST ARCHITECTURE WITH SINGLE SIGN ON METHOD ON ENHANCE SECURITY AND USER ACTIVITY MONITORING
Author:	KAMALUDIN NUR , BENFANO SOEWITO
Abstract:	The rise of cyberattacks due to weak internet access controls has led to increased data breaches and financial losses. This study proposes the integration of Zero Trust Architecture (ZTA) with Single Sign-On (SSO) to enhance network security while simplifying user authentication and monitoring. ZTA enforces strict verification for each access request, mitigating risks from phishing and brute force attacks. Despite its advantages, ZTA presents challenges, particularly in integrating with legacy systems and managing infrastructure complexity. This research evaluates ZTA implementation through network topology analysis, identity-based access control, segmentation, and Multi-Factor Authentication (MFA) to prevent unauthorized access. The integration of SSO improves authentication efficiency while maintaining security. Authentication performance is a critical factor in ZTA adoption. The results confirm that integrating ZTA with SSO facilitates real-time monitoring, enhances rapid threat response, and improves overall network security, providing a viable cybersecurity solution for modern organizations. Furthermore, this research evaluates authentication response time, aiming for an optimal range of 30-40 MS to ensure system efficiency and security. The study records an authentication time of 30.03 ms, network latency of 29.28 MS, and real-time notifications to IT staff delivered within 2.66 seconds. In live trials with a 200 Mbps bandwidth, the system detected 11 anomalies within 3 hours, demonstrating its effectiveness.
Keywords:	Zero Trust Architecture, Single Sign-On (SSO), Authentication, Network Security, User Activity Monitoring
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	THE VISUAL PERCEPTION LEARNING MODEL BASED ON INCLUSIVE DESIGN TO ENHANCE DATA STORYTELLING SKILLS IN DIGITAL VISUALIZATION
Author:	PARWAPUN KAMTAB
Abstract:	This study investigates the impact of a visual perception learning model based on inclusive design principles on improving data storytelling skills in digital visualization—an important competency in the changing landscape of information technology (IT). As data-driven decision-making becomes more prevalent across industries, effective data storytelling is critical for communicating complex insights. This study contributes to IT by introducing a systematic learning framework and interactive digital tools that assist learners in creating accessible and impactful visual narratives. The framework has five major stages: exploration, conceptualization, application, evaluation and iteration, and reflection. These stages help learners understand the relationship between visual perception and inclusive design, promoting the creation of high-quality data visualizations. Expert evaluations gave the model a high overall rating of 4.86 out of 5 stars. Participants demonstrated significant improvement in their understanding of visual perception and inclusive design, with a strong correlation between the two concepts (r=0.784). The digital learning tools, which were designed to be accessible and engaging, received positive usability feedback (M=4.73, S.D.=0.41) and effectively motivated learners. Assessment results also showed that participants improved their data storytelling skills (M=4.41, S.D.=0.64), demonstrating proficiency in both tool usage and visualization design. While the study emphasizes strengths, it also identifies challenges in simplifying complex data, creating compelling narratives, and mastering advanced visualization tools. These findings highlight the importance of inclusive design in IT education, providing learners with the necessary digital skills to create professional, engaging, and accessible visualizations. This study emphasizes the importance of continued advancements in IT-driven learning models for improving storytelling techniques and digital data visualization skills.
Keywords:	Visual Perception, Inclusive Design, Data Storytelling, Digital Visualization, Learning Model
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	AI-AUGMENTED PEDAGOGICAL FRAMEWORK FOR SUPERIOR LITERATURE REVIEW PROFICIENCY
Author:	SAWANAN DANGPRASERT
Abstract:	This study addresses the persistent challenge of inefficiency and limited precision in conducting literature reviews, which often hinder academic progress. To overcome these obstacles, the research aimed to develop and evaluate an AI-Augmented Pedagogical Framework for Superior Literature Review Proficiency, named CRAFT (Collecting, Reviewing, Analyzing, Framing, and Tailoring). Integrating advanced AI tools—such as SciSpace, Semantic Scholar, Google Scholar, Mendeley, Zotero, QuillBot, Grammarly, and Turnitin—CRAFT seeks to streamline repetitive tasks, accelerate data synthesis, and uphold academic integrity within the literature review process. A mixed-methods design was employed. Data were gathered through semi-structured interviews, online surveys, and pre-test/post-test assessments with 30 researchers from diverse disciplines. MANOVA results indicated significant improvements across five core skill areas: searching, reviewing, analyzing, synthesizing, and presenting. Participants achieved a 60.8% reduction in task completion time and attained a 92% precision rate. Qualitative thematic analysis reinforced these findings, underscoring enhanced critical thinking, deeper comprehension, and stronger interdisciplinary collaboration. Despite challenges, including initial costs for AI tools and the need for human oversight to address contextual limitations, the study concludes that CRAFT provides a flexible, accessible, and personalized framework adaptable to various research contexts. The framework holds potential to transform academic practices and promote scalable AI integration in education. Future research should focus on refining CRAFT’s adaptability, enhancing cost-effectiveness, and addressing ethical implications within diverse educational settings.
Keywords:	Pedagogical Framework, Literature Review, AI Tools Integration, Research Proficiency, Artificial intelligence
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	ADVANCES IN AI FOR PULMONARY DISEASE DIAGNOSIS USING LUNG X-RAY SCAN AND CHEST MULTI-SLICE CT SCAN
Author:	R.SRIRAMKUMAR , Dr.K.SELVAKUMAR , Dr.J.JEGAN
Abstract:	Worldwide, lung cancer, pneumonia, chronic obstructive pulmonary disease, and interstitial lung disease continue to endanger people's health. Conventional methods of diagnosing issues in the lungs often rely on chest x-rays as well as CT and ultra sound scans which need human interpretation . One area that is often taken too slowly and can at times lead to mistakes. This article looks into the most recent developments in deep learning automation systems designed for the diagnosis of chest abnormalities due to pulmonary diseases. Determining the effect of deep learning, namely convolutional neural networks, on the precision and effectiveness of clinical disease detection is the study's goal. Multiple stride deep learning models aimed at automatic recognition of pulmonary diseases will be captured with an emphasis on competition between various architectures of controlled neural nets and their algorithms. Also considered are other problems at the interface of deep learning and large medical datasets such as lack of properly labeled data, justification of model based predictions, and design of clinical decision support systems with user-hoping intelligence. Additionally, it analyzes the pioneering AI-enhanced pulmonary diagnostics and what public health outcomes can be achieved with improved timeliness and accuracy in clinical diagnosis. Afterward, we suggest critical issues that need to be addressed, such as the model verification assumptions and ethical implications regarding the use of global medical data and imaging.
Keywords:	Pulmonary diseases, deep learning, chest imaging, diagnostic accuracy, medical AI
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	NEW DISTRIBUTED ARCHITECTURE BASED ON MULTI AGENT SYSTEMS FOR THE CYBERSECURITY OF THE INTERNET OF THINGS
Author:	RACHID HDIDOU , MOHAMED EL ALAMI
Abstract:	The Internet of Things has become one of the technologies that receives great attention from researchers due to its significant impact in various fields. To secure the Internet of Things, many presented research works attempted to propose effective and suitable solutions addressing the features and characteristics of this technology. Despite all the proposed solutions, centralization remains one of the problems that affect the speed, efficiency, and accuracy of these solutions. Distribution can provide faster and more accurate solutions and avoid complete system downtime if a system’s node encounters a problem. This is our goal in this research work. In this scientific paper, a distributed architecture based on multi-agent systems has been proposed for the cybersecurity of the Internet of Things. Our architecture presents itself as an exhaustive solution as it ensures three main tasks: proactive monitoring, anomaly detection, and swift response to emerging threats.
Keywords:	Multi-Agent Systems, Cybersecurity, Internet of Things, Intrusion Detection, Distributed intelligence
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	ENGAGING MINDS THROUGH ANIMATION: THE EFFECTIVENESS OF 2D CONTENT IN MEDICAL HEALTH EDUCATION
Author:	SARNI SUHAILA RAHIM, SHAHRIL PARUMO, CHIA YEN TOH DARREN, ROSLEEN ABDUL SAMAD, SURIATI KHARTINI JALI
Abstract:	The article examines the effectiveness of 2D animated content as an engaging and innovative educational tool for mental health education. The study explores how such animations can enhance understanding and raise awareness of mental health disorders among diverse audiences. Among the ongoing difficulties mental health education faces are stigma, misunderstandings, and a dearth of interesting materials that successfully explain difficult disorders like bipolar disease. To this end, six user-friendly animated modules were developed, focusing on key aspects of Bipolar Disorder, including an introduction to the condition, its symptoms, causes, prevalence, types, and available treatments. A structured evaluation process was employed, combining usability testing to measure knowledge retention with participant feedback to capture perceptions and experiences. The study engaged three target groups: healthcare professionals, multimedia experts involved in content creation, and members of the general public with varying levels of familiarity with Bipolar Disorder. By integrating insights from these groups, the research highlights the transformative potential of 2D animation in mental health education. Preliminary findings demonstrate that animated content significantly improves comprehension and awareness, presenting it as a dynamic, engaging, and accessible medium for disseminating information. These results underscore the promise of 2D animation in reducing stigma and misconceptions about mental health disorders, paving the way for the development of future educational resources and interventions.
Keywords:	2D Animation, Multimedia Content, Educational Media, Digital Animation, Interactive Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	CHRONIC KIDNEY DISEASE DETECTION AND CLASSIFICATION USING DEEP LEARNING METHOD – AN EMPIRICAL PROOF
Author:	TEJASWINI PANSE, DR. DIVVELA SRINIVASA RAO, DR.RAGAVAMSI DAVULURI, MRS. LAXMI PAMULAPARTHY, DR.B.SENTHILKUMARAN, PONNURU ANUSHA, CH. CHANDRA MOHAN
Abstract:	A significant global health issue that is having an impact on healthcare systems is chronic kidney disease, or CKD. Prediction aids in CKD diagnosis, but the interpretability of these models remains challenging for clinicians. Explainable AI (XAI) plays a critical role in addressing this by offering intelligible insights into model predictions. By using deep learning more specifically, the VGG16 architecture instead of traditional machine learning for the identification and categorization of chronic kidney disease, this work advances earlier investigations. The aim is to enhance diagnostic precision and interpretability. Integrating insights from earlier analysis-driven studies, Grad-CAM analysis is incorporated into the deep learning framework. Providing clear and understandable representations of the neural network's decision-making process in medical image data is the aim of this integration. Medical professionals can more accurately classify and interpret images with greater intelligence when Grad-CAM analysis is integrated to highlight important areas in medical images. The discrepancy between the accuracy and interpretability of CKD diagnosis is closed by this study. Our objective is to improve the quality of healthcare worldwide by providing physicians with a useful tool for precisely identifying chronic kidney disease (CKD) by the integration of deep learning techniques and Grad-CAM insights.
Keywords:	Chronic Kidney Disease ,VGG16 ,Grad-CAM, Explainable AI (XAI), Interpretability
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	INFORMATION SYSTEM STRATEGIC PLAN ANALYSIS TO INCREASE E-COMMERCE SALES USING WARD AND PEPPARD FRAMEWORK
Author:	HENDY TANNADY , JOHANES FERNANDES ANDRY , STEFFI KRISTIANTI , ISLAMUDDIN , FURQONTI RANIDIAH
Abstract:	Analysis unit of this study was one of globally recognised e-commerce firms that operates as a consumer-to-consumer (C2C) marketplace. This firm makes the commercial center provide more buyers and provides a wider market advancement, with the help of the web the goods exchanged must be visible to anyone anytime, anywhere, no matter what, but customers can emerge from any country that is significantly different Implementing IS/IT in an organization can increase efficiency in almost all aspects, resources, business processes, markets and management. Business actors are encouraged to make strategic efforts to maintain their competitive advantage due to the rapid advancement of information technology in all fields. A SWOT analysis reveals that this firm is well positioned to take advantage of the current possibilities because of its many strengths and opportunities. This situation gives credence to plans for rapid expansion (growth-oriented strategies). The method used in this study was Ward and Peppard framework, this framework was used to analyze the IS/IT strategic plan to increase the sales number. This study concluded firm infrastructure cover across functions such as QA, financial staff and human resource management. Finance and QA personnel are involved in Enterprise Infrastructure. Shopee uses augmented reality (AR) as a means of developing new technologies. In this company, procurement is the a seller of each firm’s stores. Inbound logistics is an activity related to refunds and stock checking, then send to seller that may create opportunities for profit in the future.
Keywords:	E-Commerce, Ward and Peppard, IS/IT Strategic Plan
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	ARTIFICIAL INTELLIGENCE AND ENTREPRENEURSHIP EDUCATION: INCREASING ENTREPRENEURIAL INTENTIONS AMONG FUTURE ENTREPRENEURS IN BANDUNG CITY
Author:	WIDYA MARGARETHA , KENLY ARIESIA , OKKY RIZKIA YUSTIAN
Abstract:	The rising number of young entrepreneurs in Bandung City is now a rapidly growing phenomenon. The growth of creative industries is inseparable from the contributions of young entrepreneurs, who continually develop new innovations. However, despite this positive trend, challenges remain in effectively cultivating entrepreneurial intentions among students. One of the key challenges is ensuring that entrepreneurship education effectively enhances students' motivation and readiness to pursue entrepreneurial careers. This issue highlights the need to explore additional factors that can strengthen the impact of entrepreneurship education. This research aims to determine the effect of entrepreneurship education on the entrepreneurial intentions of business students in Bandung City and to evaluate the role of artificial intelligence (AI) as a mediating factor in this relationship. The study investigates whether integrating AI tools within entrepreneurship education can enhance students' entrepreneurial intentions by fostering creativity, decision-making, and problem-solving skills. Data were collected through surveys conducted at various universities in Bandung offering management or business programs, targeting students majoring in these fields. A total of 400 respondents participated in the survey, and the data were analyzed using Structural Equation Modeling (SEM). The results of this study show that entrepreneurship education has a significant positive effect on entrepreneurial intention. Furthermore, artificial intelligence significantly mediates the impact of entrepreneurship education on entrepreneurial intentions. These findings suggest that integrating AI tools within entrepreneurship education can enhance students' entrepreneurial mindset and motivation. This study contributes to the literature by providing empirical evidence of the mediating role of AI in entrepreneurship education. It offers valuable insights for educators and policymakers to optimize teaching strategies through the integration of emerging technologies, ultimately fostering a more innovative and entrepreneurial generation.
Keywords:	Entrepreneurship Education, Artificial Intelligence, Entrepreneurial Intention
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	DEEP LEARNING FRAMEWORK WITH OPTIMIZATIONS FOR AUTOMATIC DETECTION OF ARRHYTHMIA FROM ECG DATA
Author:	FAHMINA TARANUM , S. FOUZIA SAYEEDUNNISA , GOURI R PATIL , MANIZA HIJAB , KOTARI SRIDEVI , SYED SHABBEER AHMAD
Abstract:	The WHO states cardiovascular disorders are a significant health concern, emphasizing the need for technical advancements to provide diagnostic instruments that can identify arrhythmias or irregular heartbeats in electrocardiograms. As AI has grown in popularity, especially DL methods that have shown promise in analyzing medical data, it is imperative to apply these learning-based strategies to improve arrhythmia detection and classification performance. CD diagnosis is a promising use of current DL models, such as CNNs. Nevertheless, these models must be improved to diagnose diseases as effectively as possible. This study suggests a DL-based system for automatically identifying and categorizing electrocardiogram arrhythmias. To further apply this framework and efficiently identify arrhythmias, we provide an approach called LbADC. Our empirical investigation, which used the PhysioNet 2017 Challenge dataset as a benchmark, showed that the suggested DL architecture successfully identifies and categorizes arrhythmias in ECG data. According to the experimental data, the tested CNN model outperformed several current DL models, including LeNet, ResNet50, and U-Net, with a maximum specificity of 96.04%. Therefore, to develop a clinical decision support system for the automated screening of CD disorders, the suggested framework, the improved CNN model, and the underlying algorithm may be included in any current healthcare application.
Keywords:	Healthcare, Detection of Cardiovascular Diseases, Arrhythmias Detection, Deep Learning, Artificial Intelligence
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	HIGH-PERFORMANCE SEMANTIC SIMILARITY ANALYSIS FOR MEDICAL RESEARCH DOCUMENTS USING TRANSFORMER MODELS (BIOBERT/CLINICALBERT) WITH WMD/WMS
Author:	MAJJI VENKATA KISHORE , PRAJNA BODAPATI
Abstract:	The fast growth of medical literature poses great difficulties for finding really important and relevant research publications. While keyword-based search techniques ignore latent semantic linkages, traditional citation-based ranking systems-such as impact factors and h-index scores-often fail to adequately represent the complex influence of research. Leveraging BioBERT and ClinicalBERT models with word mover's distance or word mover's similarity, this thesis develops an advanced citation influence and semantic analysis framework that integrates parallel-influenced citation analysis with semantic similarity measures, so addressing these constraints. The methodology increases research connection by discovering semantically relevant papers that lack direct citations, bridging hidden knowledge gaps. Moreover, the suggested approach goes beyond citation counts to increase semantic similarity and relevance among several research publications detection by using deep learning-based text embeddings, thereby stressing clinically significant papers. By using this, researchers can travel beyond obsolete citation measures and identify research linked with real-time medical developments. Analyzing millions of papers with high-dimensional embeddings, however, imposes a great computational cost. To handle this, a High-Performance Computing (HPC) framework is created to parallelize similarity computations, clustering, and summarization operations. This method speeds up extensive literature review, therefore enabling real-time study discovery. This work provides a scalable, efficient, semantically enriched analysis system overall that enables researchers to find pertinent studies, rank influential publications, and more precisely and insightfully negotiate the always expanding terrain of medical literature.
Keywords:	ClinicalBERT, HPC, Semanticsimilarity, ModifiedMWD, Parrallel Computing
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	PERFORMANCE ASSESSMENT OF NHPP SOFTWARE RELIABILITY MODELS APPLYING PARETO-TYPE LIFETIME DISTRIBUTION PROPERTIES
Author:	HYO JEONG BAE
Abstract:	This study compares and evaluates the performance of NHPP-based software reliability models that apply Pareto-type lifetime distribution characteristics and proposes the optimal model based on this analysis. To analyze software failure phenomena, failure time data were utilized, and the parameters of the proposed model were determined through the maximum likelihood estimation (MLE) method. Various analyses (including assessing model efficiency using MSE and R^2, evaluating prediction accuracy against true values using the mean value function, measuring failure occurrence intensity using the intensity function, and assessing future reliability using the reliability function) demonstrated that the Lomax model exhibited the best performance. Therefore, this research provides new insights into the reliability performance of Pareto-type lifetime distributions, which have been underexplored in existing studies, and also offers fundamental reliability attribute data that software developers need in the early stages.
Keywords:	Goel-Okumoto, Lomax, NHPP, Pareto, Reliability Performance.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	TrusPass: DYNAMIC PASSWORD GENERATION FOR USER AUTHENTICATION USING GRAPHICAL KEYWORD
Author:	S. RAJARAJAN , PLK. PRIYADARSINI
Abstract:	User authentication is the gateway for entry into any system. Attacks on user authentication can cause irreperable loss for the individuals and organizations. Password based authentication is the primary method of user authentication. But they are vulnerable to many attacks. Imposing strong password policies to choose complex passwords causes difficulty in remembering them. People have the habit of using the same password across multiple accounts. This leads to cascading password attacks. Attacks on passwords commence from the moment they are being entered on the keyboard. In this paper, we have proposed a graphical keyboard based dynamic password generation scheme that facilitates inconspicuous entry of passwords. Users need to use a token for entering their passwords on the proposed keyboard. By aligning the token with the characters on the keyboard, they can unobstrusively enter their passwords. User’s password gets transformed into a high entropy dynamic password with the help of an algorithm. For the same password, different dynamic passwords are geberated each time. The scheme improves password strength without burdening users to remember complex passwords. The proposed scheme averts many attacks on passwords. The usability of the scheme is ascertained through a user study
Keywords:	User authentication, Password attacks, Internet banking, Shoulder-surfing, Form grabbing, Keylogging, Virtual keyboard, Cyber security
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	A BLOCKCHAIN-BASED MULTIMODAL APPROACH TO MALWARE DETECTION IN ANDROID IOT ECOSYSTEMS
Author:	BALINGAN SANGAMESHWAR , P. CHANDRA MOUNIKA , PRAVEENA MANDAPATI , J. MANORANJINI , TIRUMALASETTI LAKSHMI NARAYANA , V. SUJATHA LAKSHMI , D. N. V. SATYANARAYANA
Abstract:	As the Internet of Things (IoT) continues to proliferate, the security challenges associated with interconnected devices have become increasingly complex. Malicious actors exploit vulnerabilities in IoT networks, leading to the need for robust solutions in malware detection and classification. This research proposes a novel approach by integrating block chain technology into the IoT security framework to enhance the efficiency and reliability of malware detection. The study investigates the limitations of traditional malware detection methods in IoT environments and explores the potential of block chain to address these challenges. Block chain’s decentralized and tamper-resistant nature provides a secure and transparent platform for recording and validating data transactions within the IoT ecosystem. By leveraging block chain, the research aims to establish a trustworthy and resilient infrastructure for detecting and classifying malware in real-time. Furthermore, the paper discusses the implementation of a block chain-based consensus mechanism to validate the integrity of data collected from IoT devices. This consensus model ensures the authenticity of information, reducing the risk of false positives or negatives in malware detection. Additionally, the study explores the use of smart contracts to automate and enforce security policies, enhancing the overall responsiveness of the system. The proposed approach not only contributes to the advancement of IoT security but also lays the foundation for a more collaborative and secure IoT ecosystem. The findings of this research have significant implications for industries relying on IoT technologies, emphasizing the importance of proactive measures to safeguard interconnected devices from evolving cyber threats.
Keywords:	Internet of Things, malware detection, classification, block chain, IoT security
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	INTEGRATING MULTIAGENT SYSTEMS AND AIOT FOR EFFICIENT HOME ENERGY MANAGEMENT: CHALLENGES AND SOLUTIONS
Author:	A.B. PRADEEP KUMAR , K. PAVAN KUMAR , M. VENKATESWARA RAO , MUMMIDI P SUBBARAJU , V. SOWMYA DEVI , CH. LAVANYA SUSANNA , CH. RATNA BABU
Abstract:	The proliferation of IoT devices has ushered in a new era of smart homes, where efficient energy management is a paramount concern. Multiagent Artificial Intelligence-of-Things (MAIoT) has emerged as a promising approach to address the complex challenges of smart home energy management. This research paper provides an in-depth exploration of MAIoT, its key components, operation, advantages, and the challenges it faces. By leveraging the capabilities of multiagent systems in conjunction with IoT devices, MAIoT systems offer significant advantages in energy efficiency and user comfort. However, challenges such as privacy, security, interoperability, scalability, and user acceptance need to be carefully addressed. As technology continues to evolve, the future prospects for MAIoT in smart home energy management are bright, promising more intelligent and adaptive solutions to reduce energy consumption and contribute to a sustainable future.
Keywords:	Multiagent Artificial Intelligence-of-Things (MAIoT), Artificial Intelligence (AI), Smart Home Energy Management, Security.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	THE INFLUENCE OF DIGITAL LITERACY ON INTENTION TO USE INVESTMENT APPLICATIONS IN GENERATION Z: A CASE STUDY ON FINANCIAL PRODUCTS
Author:	JHONY TANIA , VIANY UTAMI TJHIN
Abstract:	Generation Z is a generation that grew up in an adequate digital environment, where technological innovation is part of their lives. Everyday Generation Z can easily receive and exchange information using applications and various other technologies. Generation Z also occupies the position as the largest population in Indonesia, this means that currently the dominant population of Indonesia is Generation Z. However, looking at the investment side, the number of investors in Indonesia is far different from the population, even though there are currently applications that support the investment process. Given that Generation Z is a digital native, this study wants to analyze the influence of digital literacy on financial attitudes that encourage someone to use investment applications to make investments. Where there are variables that are part of the study, namely: social media literacy, digital literacy, financial literacy, privacy, innovativeness, financial attitude, fintech self-efficacy and also intention to use. From this study with the results of 400 respondents, it was found that social media literacy, digital literacy, financial literacy, privacy have a significant influence on financial attitudes, and financial attitudes towards intention to use. On the other hand, the variables innovativeness and fintech self-efficacy give different results. In this study, it was found that digital literacy has a significant impact on changing a person's financial attitude to support and encourage someone to make investments. The results of this study contribute to research on investment by looking at the perspective of digital literacy as an influence that encourages someone to increase their intention to use investment applications.
Keywords:	Gen Z, Generation Z, Social Media Literacy, Digital Literacy, Financial Literacy, Investment, Intention To Use, Privacy, Innovativeness.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	AN IOT BASED EFFICIENT E-VOTING SYSTEM USING QR CODE AND EFFT-SWIFFT WITH BLOCKCHAIN FOR ENHANCED SECURITY AND TRANSPARENCY
Author:	T. PRABAKAR , S. KANCHANA
Abstract:	To modernize the democratic voting process, electronic voting systems have appeared as a promising solution. Nevertheless, concerns about the disclosure of sensitive information and the potential compromise of voter privacy might be raised by the transparency of Blockchain (BC)-centric systems. For overcoming these challenges, an Efficient E-Voting System with a QR Code using Blockchain (EVS-QCB) is proposed in this research. To strike a balance between transparency and privacy, the proposed approach incorporates an authority management mechanism. The proposed E-Voting system includes a QR code-based multi-factor authentication mechanism, unlike conventional single-factor authentication system, reducing the risk of credential theft. The system begins with the Voter Registration Phase, where the voters register to the E-Voting system. For selecting relevant attributes from voter details, the system uses a Linear Scaling-based Fox Hunting Optimization Algorithm (LS-FHOA). Next, QR codes and smart contracts are generated and stored in a secure database. Voters authenticate through usernames, passwords, and QR codes during the voting phase, and the system ensures eligibility and prevents duplicate voting utilizing Smart Contracts. Using the Public key raised to the Power of Private key-Elliptic Curve Cryptography (P3-ECC) algorithm, the voting details are encrypted. Utilizing the Exponential Fast Fourier Transformation- Single-Window Iterative Fast Fourier Transform (EFFT-SWIFFT) hashing technique, the encrypted data is transformed into a unique hash code. The generated hash codes are securely stored in BC technology. During the vote counting time, the voting details are fetched from the BC and processed for determining the final results. Performance evaluations demonstrate superior execution time, security, and memory usage compared to baseline models. This work contributes to advancing blockchain-centric e-voting security while addressing authentication and privacy challenges.
Keywords:	Linear Scaling based Fox Hunting Optimization Algorithm (LS-FHOA); Public key raised to the Power of Private Key-Elliptic Curve Cryptography (P3-
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	OPTIMIZING SENTIMENT INTERPRETATION OF COURSERA COURSE REVIEWS USING AN ADAPTIVE FISH SWARM OPTIMIZATION-INSPIRED RECURRENT NEURAL NETWORK (AFSO-RNN)
Author:	J SAHITHA BANU , G PREETHI
Abstract:	This study introduces an innovative approach, Adaptive Fish Swarm Optimization-Inspired Recurrent Neural Network (AFSO-RNN), to optimize sentiment interpretation of Coursera course reviews. The AFSO-RNN model combines the adaptive capabilities of fish swarm optimization with the power of recurrent neural networks. The recurrent neural network component processes the textual data of course reviews by capturing semantic meaning and context. It learns from sequential dependencies to extract relevant features for sentiment analysis. The adaptive fish swarm optimization component enhances the learning process of the recurrent neural network. Inspired by collective behaviour in fish swarms, it dynamically adjusts network parameters during training. The optimisation process explores and exploits optimal solutions by mimicking fish swarm movement and communication patterns, improving sentiment interpretation accuracy. Extensive experiments on a large dataset of Coursera course reviews demonstrate the superior performance of AFSO-RNN compared to traditional sentiment analysis techniques. The model’s optimized sentiment interpretation provides valuable insights into learner sentiments, enabling informed decision-making for instructors, administrators, and learners regarding course selection and improvement. This research contributes to sentiment analysis in online learning environments by showcasing the effectiveness of the AFSO-RNN model. By combining recurrent neural networks with adaptive fish swarm optimization, AFSO-RNN offers a promising avenue for enhancing the accuracy and efficiency of sentiment analysis for Coursera course reviews
Keywords:	Adaptive, Courseera, Fish Swarm optimization, Recurrent Neural Network, Sentiment Analysis, Optimization
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	EXPENDABLE DEEPFAKE DETECTION USING MULTI-STAGE DEEP LEARNING WITH SPATIAL, TEMPORAL, AND FREQUENCY FORENSIC PIPELINES
Author:	SATHYAVANI ADDANKI,MANEESHA VADDURI, VIJAYKIRAN ADDANKI, NUTHALAPATI KAMALA VIKASINI
Abstract:	Deepfake technology uses AI to create realistic but fake images, videos, and audio based on existing media. While intriguing, it poses significant threats in the digital era, affecting reputations, spreading rumors, and influencing political opinions. Advances in deepfake generation make it more convincing and accessible, increasing its misuse in cybercrimes such as identity theft, cyber extortion, fake news, financial fraud, and blackmail. To combat these threats, social media and networks seek intelligent algorithms for deepfake detection. The sophistication of deepfake technology is constantly increasing; therefore, robust and explainable deepfake detection is indispensable for digital forensics. Most existing approaches to deepfake detection focus on single-domain features, such as spatial inconsistencies, and have poor generalizability over different datasets. Moreover, they seldom handle artifacts produced by generative models like StyleGAN3, such as fine-grained blending errors and GAN-induced high-frequency noise. To overcome these limitations, we introduce an Explainable Deepfake Detection Framework that integrates a multi-stage deep learning pipeline with spatial, temporal, and frequency feature extraction. Our model begins with data collection using FaceForensics++ combined with synthetic deep fakes generated via StyleGAN3, guaranteeing diversity across compression levels, ethnicities, and poses while mitigating bias. Pre-processing employs MTCNN for face alignment and DWT for frequency domain analysis, enhancing sensitivity to subtle artifacts. Feature extraction uses three specialized modules: (1) Xception CNN for spatial features to detect blending artifacts and edge inconsistencies, (2) LSTM-based Temporal Network to capture unnatural motion artifacts over video frames, and (3) DCT-DenseNet to identify high-frequency inconsistencies in frequency space. The multi-stage fusion classifier combines the features using an Attention-Based Weighted Fusion strategy to optimize accuracy through an emphasis on influential modalities. Grad-CAM and SHAP post-processing will provide explainability by showing regions that contribute to the artifacts and quantifying the importance of features. Experiments on the FaceForensics++ dataset achieved 99.2% accuracy, 98.7% F1 score, and 0.995 AUC-ROC, making it state-of-the-art. This work not only enhances the accuracy of detection but also improves interpretability, allowing forensic experts to understand and trust deepfake predictions better in the process.
Keywords:	Deepfake Detection, Explainable AI, Multi-Stage Fusion, Forensic Analysis, GAN Artifacts
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	RECOMMENDER SYSTEMS IN E-LEARNING: TRENDS, CHALLENGES, AND FUTURE DIRECTIONS
Author:	KAOUTAR ERRAKHA ,AMINA SAMIH ,ABDERRAHIM MARZOUK , AYOUB KRARI
Abstract:	Recommender systems enable personalized learning experiences in e-learning, which was previously unheard of. This survey describes the history and evolution of recommender systems and the methodologies and problems faced in contemporary e-learning. We consider classical approaches like collaborative filtering, content-based filtering, and hybrid models, as well as new approaches using deep learning, knowledge graphs, and XAI (explainable artificial intelligence). We also analyze problems like data sparsity, cold start challenges, system interpretability, and ethics. Emerging trends like adaptive learning and proactive context-aware recommendations are also discussed. Every aspect of the field is explained in this article, and it forms the basis for further insights and the impact these systems will have on education in the future.
Keywords:	Recommender Systems ;E-Learning;Personalized Learning;Context-Aware Recommender Systems (Cars);Explainable Ai (Xai);Deep Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	FORECASTING WEEK-AHEAD CLOSING PRICE OF MUSCAT SECURITIES MARKET USING HYBRID TCN-LSTM MODEL
Author:	NASSER AL MUSALHI , MOHAMMAD NASAR
Abstract:	Accurately forecasting financial time-series data is a challenging task due to the dynamic and volatile nature of stock markets. This study introduces a hybrid Temporal Convolutional Network (TCN) and Long Short-Term Memory (LSTM) model designed to improve stock price forecasting for the Muscat Securities Market (MSM). Unlike standalone deep learning models, this hybrid approach effectively captures both short-term and long-term dependencies, leading to improved predictive accuracy. Trained on 24 years of historical MSM data (2000–2024), the model was evaluated using Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). The hybrid model outperformed both standalone architectures, achieving the lowest MAE (206.29) and RMSE (314.31). This research advances financial forecasting by introducing a hybrid TCN-LSTM model specifically optimized for the Muscat Securities Market (MSM), a relatively underexplored financial domain. The study bridges the gap in existing models by enhancing predictive performance through an innovative fusion of deep learning techniques. The study contributes to financial forecasting research by demonstrating how hybrid deep learning models can enhance market prediction accuracy, providing valuable insights for investors and financial analysts. Future research directions include the integration of adaptive learning mechanisms and external financial indicators for further performance enhancement.
Keywords:	Muscat Securities Market, Stock Price Prediction, Hybrid Tcn-Lstm, Financial Time-Series Forecasting, Deep Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	LEVERAGING HEADLESS CONTENT MANAGEMENT SYSTEM AS A SERVICE IN A SERVICE-BASED ARCHITECTURE: ENHANCING USER EXPERIENCE AND OVERCOMING RESOURCE LIMITATIONS FOR START-UPS
Author:	ILMA ARIFIANY , GEDE PUTRA KUSUMA
Abstract:	Microservices and service-oriented architecture have acquired widespread recognition in recent years as efficient solutions for building scalable, resilient, and simple-to-maintain software systems. Although they offer an optimal solution for large organizations with intricate and dynamic systems, their suitability for startups may be uncertain because of the constraints a new company may encounter, such as budget and human resources. Hence, the architectural design should be modified following the specific requirements of the company in question. This case study focused on building a service-based architecture that would address a new company’s constraints while still emphasizing the application’s user experience. We implemented this service-based architecture by utilizing a headless content management system and building additional reusable services using the plugin and database feature of the content management system. This architecture allowed developers to accelerate and simplify the development of the company’s backend services, enabling a focus on improving features impacting user experience. In our resulting architecture, each service operated on its own, with distinct responsibilities, lowering the reliance on one backend. This architecture also improved the website’s performance, as shown by a fast response time, high throughput, and an overall good load speed.
Keywords:	Composable Architecture, Headless Content Management System, Microservices Architecture, Modern Web Development, Service-Oriented Architecture
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	STRENGTHENING SECURITY PROTOCOL TO COMBAT FINANCIAL FRAUD ADVANCES IN AUTHENTICATION AND ACCESS CONTROL
Author:	JANJHYAM VENKATA NAGA RAMESH , DR.M.V.RAJESH , B.VEERAJYOTHI , VEMULA JASMINE SOWMYA , AMIT VERMA ,REFKA GHODHBANI
Abstract:	This research delves into the potential of combining machine learning with multi-factor authentication (MFA) to strengthen financial system security protocols. The main objective is to fight fraud using improved authentication and access control methods. The study looks at practical implementation techniques to fix current security flaws, such as issues with centralised authentication systems that are vulnerable to hacks and system malfunctions. The study highlights the value of energy-efficient and ecologically friendly security measures in addition to traditional encryption and biometric techniques. Different architectures, consensus protocols, applications, services, and implementation goals are all covered in the analysis of various security frameworks. Six machine learning models are thoroughly examined to see how well they identify fraudulent activity and improve overall money security. The results demonstrate how ML and MFA together may greatly enhance network intrusion detection and fraud prevention, eventually improving the security of financial transactions.
Keywords:	Machine Learning, Network Intrusion Detection, Multi-Factor Authentication Security Frameworks
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	SECURE AND RESILIENT VIDEO BLOCK MANAGEMENT AND PRINCIPLES FOR ZERO TRUST CLOUD NETWORK
Author:	K.MUTHULAKSHMI, K.VALARMATHI, J.SATHYAPRIYA, S.PRIYADHARSHINI
Abstract:	Cloud computing principles are widely applied to provide reasonable network services to a range of users. In addition to the basic services of cloud networks, security features are highly expected at the user end. Considering these issues as a scope of this article, various existing techniques are identified. The existing solutions on zero trust cloud network target single mode encryption model irrespective of data models. However, the video content delivery systems in cloud platforms need multiple modes of secure data transmissions. According to that, the proposed Secure and Resilient Video Block Management (SRVM) model has been developed to attain both real-time and non-real-time security constraints. On the basis, this proposed model uses switchable block cipher and stream cipher model with respect to video data transmission modes. Similarly, the data blocks collected in the cloud environment are effectively distributed among cloud virtual machines and data is encrypted in a complex manner. This kind of novel idea reduces the chances of data confidentiality breaches and data losses due to attackers. The experiments have been conducted between SRVM, Blockchain based Multimedia Data Security (BMDS), Firefly Optimization and Encryption (FFOE), and Quality of Everything in Edge based Encryption (QMEE) to ensure the testbed performances. In this experiment, the proposed model works 10% to 15% optimally than existing techniques and the details are observed through multiple test cases.
Keywords:	Cloud Computing, Security, Video Data Management, Virtual Machines And Data Confidentiality.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	OPTIMIZING DEEP LEARNING MODELS FOR ACCURATE MULTICLASS BRAIN TUMOR CLASSIFICATION IN MRI IMAGES
Author:	GOKAPAY DILIP KUMAR , PREM SWARUP MALLIPUDI , SHAIK THASEENTAJ , POLISETTY SWETHA , N RAMA RAO , SRI KUMARAN R P , MUDIYALA APARNA
Abstract:	A brain tumour is an aberrant cell mass that has the potential to become cancerous. Magnetic resonance imaging (MRI) scans are frequently used to identify brain tumours. An MRI can provide information about the abnormal growth of brain tissue. Deep learning is used to construct models for the identification and categorization of brain tumours using MRI. This makes it easier and faster to diagnose brain tumours. Brain tumours can be found more quickly and precisely by using these algorithms to analyse magnetic resonance imaging (MRI) scans, which will help patients receive better care. If brain tumours were consistently detected early and treated appropriately, the death rate might be reduced. This paper also discusses the proposed design and compares it with other models, such as DenseNet-169 and Inception-ResNet-v2. By applying the recommended models, we achieved accuracy levels of 97.52% and 96.89%. In contrast to other models, it yields the best results.
Keywords:	MRI, Deep Learning, Brain Tumor, Health Care.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text

Title:	PERSONALIZED SLEEP SIGNATURE: A NOVEL APPROACH TO UNVEIL PAEDIATRIC SLEEP BEHAVIOUR WITH TRANSFORMER ATTENTION MECHANISM AND GRAPH ATTENTION NETWORKS
Author:	AMALA ANN K A , DR. VAIDEHI V
Abstract:	Sleep disorders manifest differently across individuals, making accurate diagnosis and treatment highly complex. Even within the same diagnosis, there can be variation in sleep architecture among patients, which makes generalization across people difficult. Traditional sleep analysis methods rely on manual scoring and fixed diagnostic criteria, which fail to capture subject-specific variability in sleep patterns. To address this, we propose a data-driven Personalized Sleep Signature (PSS) approach that learns individualized sleep behaviour using AI models. This study introduces the PSS framework, combining Transformer-based Attention and Graph Attention Networks (GATs) to model nuanced sleep characteristics. We utilize the Nationwide Children’s Hospital Sleep Data, a paediatric Polysomnography (PSG) dataset containing EEG and physiological parameters such as ocular movements, EMG activity, blood pressure, and respiratory rate. From this, we extract sleep epoch features and demographics to form Sleep Signature Groups that reflect common behavioural patterns. Unlike conventional classification, our method captures personal variability and delivers individualized sleep hygiene guidance. The model achieved 94% accuracy in detecting sleep patterns, outperforming traditional methods. Beyond clinical applications, it can be integrated with wearable sensors (e.g., Fitbit, Oura, Apple Watch) to personalize wake/sleep routines and environments. It also enables early detection of sleep disorders and aligns daily schedules with individual chronotypes to enhance well-being. By focusing on sleep behaviour rather than rigid diagnostic categories, this approach supports non-pharmacological, personalized interventions backed by scientific evidence. Our work opens the door to precision sleep medicine, offering actionable insights for clinicians, researchers, and technology innovators.
Keywords:	Personalized Sleep Signature, Transformers, GAT, NCH Sleep Data, Polysomnography
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2025 -- Vol. 103. No. 7-- 2025
Full Text