		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
May 2025 | Vol. 103 No.9

Title:	MULTI-CNN MODEL TO EVALUATE THE PERFORMANCE OF FACE DETECTION AND RECOGNITION WITH FACIAL FEATURE DETECTION AND RECOGNITION
Author:	G. SUJATHA, MATTA SWATHI, BHAGYA PRASAD BUGGE, SHAIK JOHNY BASHA, ALLURI SWATHI, BALA PRASANTHI PAVULURI, M SITHA RAM, SURYA PRASADA RAO BORRA
Abstract:	Face Recognition is one of the most advanced and drastically growing research areas because it helps identify people globally in various ethical and unethical applications. Face recognition needs face detection that can be compared with a list of available faces to predict the correct person. Face detection has become popular, easy, and fast since it follows the Viola-Jones FD method. Face comparison is obtained by comparing the internal and external information from the face images, like different features, face structure, key points, and patch-by-patch comparison. Earlier face recognition methods used separate algorithms for feature extraction from the face images, like color, shape, texture, histogram, and local and global binary pattern, to compare pairs of images where they provide more complexity regarding computation, cost, and time. After the evolution of artificial intelligence models, recent research has focused on using machine and deep learning algorithms for face detection and recognition. However, the accuracy of face recognition models needs to be improved under various conditions. Thus, this paper used a two-stage face comparison model to enhance face recognition efficiency. A consequence of three CNN models called CNN-1, CNN-2, and CNN-3 are used to detect the faces, detect the facial features, and recognize the faces, respectively. The CNN models are implemented in Python, and the results are verified by experimenting with multiple benchmark face datasets. The output accuracy obtained from the face detection and recognition is compared with the facial feature detection and recognition to choose the best to identify the criminals. From the comparison, both FDR and FFDR obtained 99.68% accuracy equally
Keywords:	CNN, Deep Learning Algorithm, Face Detection, Facial Feature Detection, Face Recognition.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	NOVEL ONE PROTOTYPE IN EACH OF THE CLASSES EMBEDDED IN FUZZY K-NEAREST NEIGHBOR
Author:	PAYUNGSAK KASEMSUMRAN , TATSANEE CHAIYA
Abstract:	Many researchers in the world are interested in classification algorithms. One of the most famous and popular algorithms is K-NN and Fuzzy K-NN. Nevertheless, the time complexity of both algorithms is . In our previously publication, we introduced the novel algorithm in which each of the classes has one representation. Although it is a good algorithm and helps to solve a time process problem but there are still problems with the accuracy rate of classification. Thus, we develop a novel one prototype in each of classes embedded in Fuzzy K-Nearest Neighbor to solve the problem for accuracy rate of classification. The system provides 99.55%, 98.25%, 83.43%, 69.07%, 80.01%, 100%, 96.25%, 81.81%, 80.65% and 83.27%% in MIT-CBCL, ORL, FEI, Georgia Tech, Pain Expression, JAFFE, Senthikumar, Yale, PICS and CMU AMP databases, respectively.
Keywords:	Prototype, Embedded, Fuzzy K-NN, Representation, Novel algorithm, NsgFK-NN1P.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	AN INNOVATIVE MODEL OF ARTIFICIAL INTELLIGENCE-BASED DOCTOR ASSISTANT MODEL FOR ADVANCED AND INTELLIGENT MEDICAL SUPPORT SYSTEMS
Author:	S. SЕLVARAJ, S. RAJAPRAKASH , SHANTHA SHALINI K , Dr R. MARIAPPAN
Abstract:	By improving patient care, operational effectiveness, and medical decision-making, artificial intelligence (AI) is transforming the healthcare industry. This study examines the idea of doctor assistants driven by AI, emphasising how intelligent medical support could revolutionise healthcare delivery. We offer a thorough examination of AI's uses in clinical decision-making, patient monitoring, treatment planning, and medical diagnosis. We also evaluate current AI-powered doctor assistant systems, pointing out their benefits and drawbacks.This study's main contribution is a comprehensive assessment of AI medical assistants that incorporates the technological, moral, and practical issues that are frequently examined separately. Our research offers a multifaceted framework that takes into account data quality, trust, explainability, clinical workflow integration, and regulatory compliance, in contrast to previous studies that just concentrate on AI accuracy or ethical issues. Along with addressing resistance, scepticism, and the need for ongoing learning, this study offers a series of recommendations for enhancing AI adoption in actual healthcare settings. This study promotes the responsible implementation and ongoing development of AI doctor assistants in contemporary healthcare by addressing these issues and assessing current developments. In addition, we suggest future lines of inquiry to guarantee the efficacy, openness, and moral rectitude of AI-powered medical devices.
Keywords:	Artificial Intelligence, Doctor Assistants, Intelligent Medical Support, Machine Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	PREDICTION OF GEOMETRIC FEATURES IN 3D FACIAL SOFT TISSUE IMAGES USING FULLY CONNECTED NEURAL NETWORK
Author:	CHANDRA SEKHAR KOPPIREDDY , SIVA NAGESWARA RAO
Abstract:	The relevant geometric prominence of the three-dimensional facial soft tissue images is a crucial factor in forensic reconstruction, craniofacial surgery, and biometric identification. The image processing technique is one of the most critical steps in analysing 3D facial images. Traditional image processing methods often face problems while capturing the soft tissue image's non-linear deformations and complex anatomical variation. So, this paper presented a solution to this challenge by formulating an accurate geometric feature prediction model using Fully Connected Neural Network (FCNN) for 3D soft tissue facial images. The important feature of the model is that it was trained on high-definition 3D face scans and leveraged deep learning-driven feature extraction to increase the model's accuracy. Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Structural Similarity Index (SSIM) are the metrics used for evaluation. The presented model performed better than traditional deep learning models based on the results. Finally, the proposed model has provided far more accurate results in shape prediction, which makes this model highly suitable for application in personalized facial reconstruction, clinical diagnostics, and aesthetic surgery planning. These enhancements in the technique contribute to the development of an improved automated 3D facial analysis, making a better way of modelling the soft tissue area and the computational facial anatomy.
Keywords:	3D Facial Soft Tissue, Geometric Feature Prediction, Fully Connected Neural Network, Deep Learning, Craniofacial Analysis.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	THE STUDY OF OPTIMIZATION OF THE SEGMENTATION PROCESS FOR BUILDING DISTRIBUTED GENERALIZED SUFFIX TREES
Author:	IAN SADOVYI , VOLODYMYR VOROTNIKOV
Abstract:	Suffix trees are important data structures used for substring search and analysis of large text sequences, but building them on large data volumes remains a challenging task. Existing methods for constructing distributed generalised suffix trees (GSTDs) have significant limitations in terms of segmentation efficiency under resource constraints. Previous studies have focused on static memory allocation strategies that do not provide an optimal balance between performance and resource consumption. Therefore, the aim of the study was to develop new segmentation methods for efficient data distribution between nodes of distributed systems. The methods with the following algorithms were compared for this purpose: hybrid division with load balancing, fractal division and dynamic adaptive division, as well as their integration with the Master-Worker architecture. The research employed the methods of modelling, experimental method, comparative method, and statistical analysis. The results showed that Master-Worker integration significantly improved performance, including reduced execution time, improved speedup and efficiency, and improved segment merging accuracy. The highest indicators were achieved when applying dynamic adaptive division, which turned out to be the most effective for data sets of different structures. Experimental results demonstrate a 30% reduction in computational costs compared to traditional methods. Thus, this study makes a significant contribution to improving algorithms for building distributed suffix trees and their application in big data analysis. Further research may focus on optimizing segmentation methods using machine learning (ML) to automatically select a strategy based on the data type. Besides, it is promising to extend the proposed algorithms to other types of data, such as multimedia streams and graphs.
Keywords:	Optimization, Suffix Tree, Distributed Computing, Parallel Computing, Data, Algorithms, Segmentation.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	DIGITAL PLATFORMS FOR SOCIAL SERVICES ADMINISTRATION IN THE CONTEXT OF SUSTAINABLE COMMUNITY DEVELOPMENT
Author:	VITALII SEROHIN , SVITLANA SEROHINA , LIUDMYLA HORBATA , PETRO GUDZ , OLHA MALTSEVA
Abstract:	The relevance of using digital platforms for social services administration is rapidly growing in the context of the digitalization of society and the need to ensure sustainable community development. The aim of the article was to analyse the effectiveness of using digital platforms for managing social services in the context of sustainable community development in Europe. The research employed the following methods: structural equation modelling (SEM) to analyse the relationships between platform performance indicators, the diffusion of innovations theory to assess the speed of their implementation, as well as comparative and content analysis to arrange data. The t-test was used to check statistical significance, the platform productivity function, as well as correlation analysis to identify relationships between the Social Progress Index (SPI) and the Sustainable Development Goals (SDGs). The results of the study show that the integration of digital platforms contributes to increasing the accessibility and efficiency of social services. A trend towards reducing the administrative burden and accelerating the achievement of sustainable development goals is identified. The article examines technical limitations and uneven access to digital technologies in European regions. The proposed ways of improvement include the development of national digital strategies, the integration of modern blockchain-based tools. The practical significance of the study is the provided recommendations for improving digital mechanisms of social governance based on the experience of European countries. Further research should be aimed at studying global practices and developing universal solutions for the digitalization of social services.
Keywords:	Sustainable Development, Territorial Communities, Sustainable Development Goals, Public Administration, Information Society, Digitalization.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	INTERDISCIPLINARY VIRTUAL LEARNING COMMUNITY MODEL FOR SOCIAL ENGINEER
Author:	THANANAN AREEPONG , PRACHYANUN NILSOOK , PANITA WANNAPIROON
Abstract:	The paper suggests developing a Metaverse interdisciplinary learning community model - M-ILC - to develop social engineers. The interdisciplinary learning community process leverages the Metaverse platform tool to foster social engineering skills among students. The study offers a synthesis of materials with regard to interdisciplinary learning communities in various formats. Emphasis is placed on the significance of nurturing human soft skills through utilizing the Metaverse in the learning process to provide learners with a 3D virtual experience. This collaborative learning approach leads to a more profound comprehension of subject content and expands educational opportunities for students. The suitability of the Metaverse Interdisciplinary Learning Community model (M-ILC) developed by experts in Information Technology, Communication Technology, and the Metaverse, was assessed. The evaluation results were rated as "excellent", indicating the suitability of the overall learning community model. This suggests that the M-ILC model can effectively cultivate students' social engineering skills and prepare them for the upcoming digital transformation. Furthermore, it contributes to sustaining a consistent quality standard in the education system. The researchers have introduced new teaching concepts and innovations that align with the current situation in the form of a learning model, fostering a boundary-less learning society that can be accessed anytime and anywhere.
Keywords:	Metaverse, Interdisciplinary, Social Engineer, Virtual Learning Community
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	REVOLUTIONIZING HEALTHCARE WITH LARGE LANGUAGE MODELS: ADVANCEMENTS, CHALLENGES, AND FUTURE PROSPECTS IN AI-DRIVEN DIAGNOSTICS AND DECISION SUPPORT
Author:	N MOHANA PRIYA1, AVANI ALLA , S PHANI PRAVEEN , YADAIAH BALAGONI , NARASIMHA RAO TIRUMALASETTI , VAHIDUDDIN SHARIFF , U GANESH NAIDU
Abstract:	Large Language Models (LLMs) such as BERT, GPT-3, and GPT-4 play a central role in revolutionizing diagnostic support and healthcare decision-making. These premium AI solutions essentially boost machine clinical analysis through vast datasets such as electronic health records (EHRs), medical literature, imaging data, and genomic data. Deep learning and NLP features in LLMs allow for disease detection at an early stage, treatment plans as per the patient as well as monitoring of the patient; all these make clinical procedures easier. However, virtual assistants powered by AI as well as chatbots keep patients interactive while easing the burden of healthcare professionals. Of course, with all the benefits LLMs have contributed to the world of today-processing much work in bulk at fast speed-globally they raise critical challenges such as data security concerns bias in AI system predictions and transparency in decision-making. Ethical use of AI in healthcare demands adherence to legislation like HIPAA and GDPR to protect data with principles based on fairness accountability and interpretability. IMMultimodal integration of textual visual genetic data will be more precise but requires federated learning transfer learning and data augmentation such that models are enhanced without compromising privacy. Cooperative effort among AI researchers, physicians who give input on policy development towards developing trusted transparent fair standards-based diagnostic tools with the help of ML in medicine is necessary. The present study explores how LLMs have evolved over time in the healthcare industry in terms of challenges encountered as well as probable outcomes integrated into democratizing medical information along with progress in precision medicine.
Keywords:	Large Language Models (LLMs), AI in Healthcare, Predictive Diagnostics, Medical Decision Support, Electronic Health Records (EHRs), AI Ethics in Medicine, Multimodal AI Systems and Federated Learning in Healthcare.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	PARKINSON'S DISEASE AND AUTISM SPECTRUM DISORDER DIAGNOSIS USING SIGNIFICANT FEATURES AND ENSEMBLE DEEP LEARNING
Author:	DR. A. SYED HAROON , DR JAISINGH W , Dr. KAARTHIEKHEYAN V, Dr. LEELAMBIKA KV
Abstract:	More than 6 million people globally suffer from progressive Neurological Disorders (NDD) such as Parkinson's disease (PD). Furthermore, a variety of developmental impairment disorders that have an impact on a patient's capacity for social interaction and communication are collectively referred to as Autism Spectrum Disorder (ASD). Conventional methods for detecting PD and ASD, however, are frequently manual and need expertise. To solve this issue, a prior study presented an Ensemble Feature Selection (EFS) method based on the Fuzzy-based Beetle Swarm Optimization Algorithm (FBSOA) and Score-based Artificial Fish Swarm Algorithm (SAFSA). This technique chooses the most significant features from the database to increase the PD detection rate. First, Min-Max Normalization (MMN) was used to normalize the dataset's scale. Dimensionality reduction (DR) is achieved by the application of Binomial Cumulative Distribution Function-based Principal Component Analysis (BCDPCA). The optimal features for PD classification are then found via Ensemble-based FS (EFS). To classify PD, Ensemble Learning (EL) classification approaches such as Fuzzy K-Nearest Neighbor (FKNN) hybrid classifier, Kernel Support Vector Machines (KSVM), Fuzzy Convolutional (NN) Neural Network (FCNN), and RF are employed. However, dataset which are used for this work has missing values which will leads to the misclassification error. And also, MMN does not handle outliers very well. Resolving those challenges, this study proposed an improved model for PD and ASD detection. Then first missing values are imputed using Most Common value (MCI). And data normalization will be done by ZSN. DR will be performed based on BCDPCA. Significant features are selected from the dimension reduced data using ensemble of Enhanced Chicken Swarm Optimization (ECSO) and Improved Whale Optimization (IWO). Finally, PD and ASD are detected using ensemble of FCNN and LSTM network. The suggested schema obtained better performances for obtained values of precision, recall, and f measure in experimental outcomes.
Keywords:	Parkinsons Disease (PD), Autism Spectrum Disorders (ASD), Detection, Z Score Normalization (ZSN), missing values, Long Short-Term Memory (LSTM) network, Chicken Swarm Optimization (CSO), Whale Optimization (WO).
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	SMART PRICING SOLUTIONS FOR USED CARS USING DECISION TREE AND RANDOM FOREST MODEL
Author:	PRATAP KUMAR CHAMUPATY , UDITA J. MONANI , BISWAJIT DAS , PRASANT KUMAR PATTNAIK
Abstract:	The used car market is vast and is affected by many parameters, making it is essential to predict prices of used cars. This work recommends customers in making informed decisions while planning to purchase used car. Presently, Machine Learning emerged as the most effective methods for prediction of prices of used cars considering correlated factors such as mileage and vehicle age. This work applied two machine learning models includes Decision Tree and Random Forests on large used car dataset so collected from Kaggle. The prediction accuracy and precision of these models thoroughly evaluated. The comparison of results output of two models is made to find out the effectiveness and reliability for analyzing large database. The study concludes that Random forest model emerged as the best model in comparison to Decision Tree model in prediction accuracy.
Keywords:	Machine learning; random forest; decision tree; used/pre owned cars; regression; predictions
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	SEAL OPTIMIZATION DRIVEN RANDOM FOREST FRAMEWORK FOR ENHANCED ALZHEIMER’S DISEASE CLASSIFICATION
Author:	V. HARIPRIYA , P. RUTRAVIGNESHWARAN
Abstract:	Alzheimer s Disease remains a progressive neurodegenerative disorder, necessitating early detection for effective intervention. Traditional diagnostic methods face challenges in accuracy, feature selection, and handling high-dimensional neuroimaging data. The purpose of this study is to enhance classification performance by integrating bio-inspired optimization with Machine Learning. A Seal Optimization-driven Random Forest (SO-RF) framework is introduced to refine feature selection, optimize hyperparameters, and improve decision-making in disease classification. The methodology involves leveraging seal-inspired search strategies to enhance the diversity and robustness of the Random Forest model, ensuring balanced precision and recall. The proposed SO-RF model outperforms conventional approaches in classification accuracy, sensitivity, and specificity, demonstrating its effectiveness in reducing false positives and false negatives. Experimental results validate the model’s superiority in handling complex medical data, confirming its potential for automated Alzheimer’s Disease diagnosis. The optimized classification framework presents a promising solution for advancing computational techniques in neurodegenerative disease detection.
Keywords:	Alzheimers disease Classification, Bio-Inspired Optimization, Seal Optimization, Random Forest, Feature Selection, Healthcare.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	FOCUSNET-LC: A DEEP LEARNING FRAMEWORK AND ALGORITHM FOR EFFICIENT AND EXPLAINABLE LUNG CANCER DETECTION
Author:	S. SUDESHNA DR. B. UMAMAHESWARA RAO
Abstract:	For lung cancer, one of the leading global causes of cancer death, accurate and early detection is essential to improve treatment success. Current practices for classification response prediction, such as DL-LCD and MFDNN, achieve high accuracy but are limited by the fact that they (i) lack interpretability, (ii) require multimodal data, and (ii) are inefficient in concentrating on clinically relevant areas. These limitations prevent their implementation in real clinical contexts and highlight the need for innovative and reliable interpretative solutions. This study introduces the FocusNet-LC model and FocusNet-LC Based Lung Cancer Detection Algorithm, which is built to deal with these difficulties. The FocusNet-LC model combines region of interest (ROI) segmentation, metadata incorporation, and model explainability (Grad-CAM) to deliver accurate, interpretable, and clinically relevant predictions. The corresponding algorithm utilizes the features from the model to efficiently process data, extract features, and classify them. The proposed framework achieves 97.36% accuracy on the IQ-OTH/NCCD dataset and produces higher precision, recall, and F1-score compared to state-of-the-art models. Thus, the usefulness of this framework comes from providing explanatory and correct predictions with calculation speed. This study also represents a significant advancement in early lung cancer diagnostics and improved patient management by not just filling in existing methodological gaps.
Keywords:	FocusNet-LC, Lung Cancer Detection, Deep Learning, Region of Interest (ROI) Segmentation, Explainable Artificial Intelligence (Grad-CAM)
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	MACHINE LEARNING BASED PERSPECTIVES FOR HOUSING MARKET CRASH PREDICTION
Author:	SIPRA SAHOO , MADHURI RAO , SMITA RATH , DEEPAK KUMAR PATEL , MITRABINDA KHUNTIA , SHRABANEE SWAGATIKA ,SUSHREE SANGITA JENA , PRABHAT KUMAR SAHU
Abstract:	The recession of 2008, commonly referred to as the "Subprime Mortgage Crisis," was the worst to hit the USA and had an impact on the entire world, costing more than 8.4 million people their employment alone in the USA. From late 2006 to the end of 2009, the crises repeatedly forced innocent people to leave their homes, and the Gross Domestic Product (GDP) dropped as low as 4.3%. In order to assist in estimating future market values and prices so that a real estate market recession can be averted, this working paper analyses old and present prices, levels, and interest rates to determine at what rate the market could have been rescued from falling. A house market crash prediction with machine learning techniques including Linear Regression, Hidden Markov Model, and Long Short Term Memory is presented here.
Keywords:	Hidden Markov Model; Linear Regression, Long Short-Term Memory; Mortgage-Backed Securities (MBS); House Market Crash
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	ROBUST AND EFFICIENT SUPPLY CHAIN MANAGEMENT USING IMPROVED PROOF OF USEFUL WORK (POUW) CONSENSUS ALGORITHM
Author:	KOLLI LALITHA KUMARI, Dr.P. LALITHA SURYA KUMARI
Abstract:	Recently, blockchain technology has been integrated with other emerging technologies, including data science, artificial intelligence, and the Internet of Things (IoT). Blockchain will be the best alternative for maintaining data security and transparency, enhancing network management. This paper presents an improved consensus algorithm, an extension of the existing PoUW, to address scalability issues in blockchain technology. The proposed consensus algorithm includes two additional modules, authorisation and storage availability, to maintain the scalability of the blockchain network. The proposed method is applied explicitly to the supply chain management framework to analyse how it overcomes the limitations of the traditional system. Measuring the supply chain framework using blockchain is facilitated by several key parameters, including latency, transaction throughput, and computational energy. Authorised people can only access and update the chain details. Based on the storage availability, it adds a new block to the blockchain. The primary objective of the proposed methodology is to enhance network scalability and improve blockchain usability in real-time applications, such as supply chains, thereby reducing the risk of counterfeiting. From manufacturer to end-user, introducing blockchain technology will help maintain transparency among all parties, including manufacturers, distributors, retailers, and customers. The proposed algorithm provides security to the network and confidentiality to sensitive information.
Keywords:	Consensus Algorithm, Authorization, Storage Availability, Supply Chain Management
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	SECURE AND SCALABLE BLOCK CHAIN-INTEGRATED AI MODEL FOR PLANT DISEASE DETECTION
Author:	ANURADHA ANUMOLU, DR. SHAHEDA AKTHAR
Abstract:	The occurrence of crop diseases creates substantial danger for both agricultural production outputs and system stability. Accurate and early detection is essential for mitigating crop losses, but existing AI-based methods often suffer from challenges in noise sensitivity, data integrity, and computational inefficiency. This paper proposes an advanced plant leaf disease detection system integrating Vision Transformers (ViT) for feature extraction, Reinforcement Learning (RL) for feature optimization, and block chain technology for secure and decentralized data management. Experimental results on the Plant Village dataset demonstrate a 97.8% accuracy with a reduced processing time of 68.2 seconds. Block chain integration further ensures data transparency and immutability, setting a new benchmark for scalable and trustworthy plant leaf disease detection.
Keywords:	Plant Disease Detection, Block chain, Reinforcement Learning, Vision Transformers, Agricultural AI.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	PERFORMANCE ANALYSIS OF THF: A NOVEL LIGHTWEIGHT CRYPTOGRAPHY HASH FUNCTION
Author:	MANGALAMPALLI K S MANYAM , KUNJAM NAGESWARA RAO
Abstract:	Cryptographic hash functions play a crucial role in contemporary information security systems, serving as foundational elements for ensuring data integrity and authentication in a wide array of applications. With the ever-growing demand for secure communication and robust data protection, there is a pressing need for hash functions that not only provide strong security guarantees but also operate efficiently in environments with limited computational and power resources. Existing cryptographic hash functions, such as SHA-2 and SHA-3, offer high security but often impose large computational overhead, making them unsuitable for resource-constrained environments. In addition, lightweight hash functions proposed in recent studies either compromise on security or fail to achieve optimal efficiency, leaving a gap in the development of balanced solutions. Recognizing this need, the authors previously introduced a novel lightweight cryptographic hash function, termed the Tiny Hash Function (THF), specifically designed to meet these constraints. This research article delves into an extensive performance analysis of THF, examining its efficiency, security, and suitability for use in resource-constrained settings. The study systematically evaluates THF in comparison to existing lightweight and standard hash functions, addressing critical aspects such as energy consumption, computational complexity, and resistance to cryptographic attacks. The analysis aims to validate THF's effectiveness in providing the desired security properties while maintaining minimal resource consumption, thereby making it an attractive choice for applications such as Internet of Things (IoT) devices, mobile platforms, and other scenarios where traditional hash functions may be impractical. By bridging the gap between security and efficiency, this study contributes to the advancement of cryptographic primitives tailored for modern lightweight computing environments.
Keywords:	Cryptographic Hash Functions, Information Security, Lightweight Cryptography, Tiny Hash Function (THF), Efficiency.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	A GAN-BASED METHOD TO TUNE LSTM HYPERPARAMETERS FOR FINANCIAL FORECASTING
Author:	ADNANE EL OUARDI , BRAHIM ER-RAHA , MUSTAPHA RIAD , KHALID TATANE
Abstract:	Optimizing hyperparameters is a critical challenge in enhancing the performance of Long Short-Term Memory networks for financial time series forecasting. Traditional optimization techniques such as grid search and random search are often computationally expensive and inefficient, while Bayesian optimization, despite its advantages, can struggle with exploration in complex search spaces. This paper introduces a novel Generative Adversarial Network-based approach to LSTM hyperparameter optimization, specifically applied to forecasting the next closing price of the S&P 500 index. The proposed method consists of a generator, which suggests potential hyperparameter configurations, and a discriminator, which evaluates their effectiveness based on forecasting accuracy. Through iterative adversarial training, the generator refines its suggestions, dynamically adapting to the optimization landscape and effectively balancing exploration and exploitation. The performance of the GAN-based optimization approach is evaluated using metrics such as Mean Squared Error, execution time, and resource utilization. Experimental results demonstrate that the proposed approach achieves competitive accuracy while improving efficiency and robustness in navigating the hyperparameter space. The findings of this study provide valuable insights into the application of adversarial learning for hyperparameter tuning, offering a promising alternative for enhancing LSTM-based financial forecasting models, particularly for the S&P 500 index.
Keywords:	Hyperparameter Optimization, LSTM Networks, Generative Adversarial Networks, Time Series Forecasting, Machine learning in finance
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	PREDICTING STUDENT ACADEMIC PERFORMANCE USING ENGAGEMENT FEATURES: A PROCESS MINING AND DEEP LEARNING APPROACH
Author:	DEENA M. HELAL , YEHIA M. HELMY , DOAA S. ELZANFALY
Abstract:	In the digital era, the increasing availability of data from online educational environments enables advanced analysis and prediction of student academic performance. As a key indicator of student progress and achievement, academic performance necessitates effective tools for analysis and intervention to enhance learning outcomes. This study integrates process mining, deep learning to predict academic performance with 99.86% accuracy for intermediate grades and 92.48% for final scores, using engagement features like mouse clicks and keyboard strokes from a widely recognized dataset spanning six sessions. Through novel feature extraction and various preprocessing techniques applied with process mining and deep learning approach , we identify that engagement behavior significantly correlate with academic success. The findings confirm the predictive strength of engagement features, providing actionable insights into student interactions and learning behaviors to inform targeted interventions.
Keywords:	Process Mining , Deep Learning , Machine Learning , Predicting Academic Performance.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	AN EMPIRICAL STUDY OF THE SIX SIGMA APPROACH FOR REDUCING THE NUMBER OF COBBLES - STATISTICAL SOFTWARE QUALITY ASSURANCE TECHNIQUE
Author:	GEETHA GUTHIKONDA, MADHURI NALLAMOTHU, VENU. ALURI, EERLA RAJESH, DR. KOLLURU SURESH BABU, DR SURESH BETAM, DR. CHANDANAPALLI SURESH BABU, CH. CHANDRA MOHAN
Abstract:	The blooms of size 250320sqmm size which are being produced at steel melt shop are kept at bloom storage yard. Accordingly, these blooms are charged into the furnaces and rolled into the billets of size 125125sqmm. The bloom is converted into billets when drawn through seven stands of variable speeds. Later the billet is passed through the shear to get cut into three pieces for passing to the Wire Rod Mill (WRM) feeding. The material passing from stand-1 is continuous and passes through the stand-7 where it gets the cross section of 125*125sqmm. whenever a problem arises due power failure or any other malfunctioning; the material struck up and is twisted in between the stands which is known as COBBLE. Removing the cobble takes minimum three to four hours which hampers the production and is a direct loss to the company. Thus the problem needs to be looked into and the solution aroused in the form of Six Sigma tools. Six Sigma is widely recognized as a business that employs statistical and non-statistical tools and techniques, change management tools, project management skills, team work skills and a powerful roadmap (DMAIC) to maximize an organization’s return on investment (ROI) through the elimination of defects in processes. DMAIC methodology provides with a data-driven methodology for achieving sustained process improvements by reducing defects. This methodology is used to identify inadequacies in the existing process and to make lasting and controllable changes to products and processes that improve product quality, customer satisfaction and the company’s profitability. Thus, with the help of proper tools and techniques of this procedure the problem had been handled. Sufficient data was collected and the various factors that affect the process and lead to failure were analyzed. The Six Sigma roadmap enabled to identify the potential causes of the failure and provided effective methods to eliminate them or minimize them. The project is all about the way the problem was tackled and the process was improved.
Keywords:	Six Sigma, DMAIC, Cobble, Breakdown mill, WRM
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	HOW CAN ARTIFICIAL INTELLIGENCE SHAPE THE FUTURE OF SUSTAINABLE EDUCATION? CHALLENGES AND OPPORTUNITIES
Author:	NISREAN JABER THALJI , SANI ALKHASAWNEH
Abstract:	Artificial intelligence (AI) is bringing about a significant transformation in the educational process, making it more personalized for each student and enhancing their engagement with content while reducing the time spent on administrative tasks such as attendance tracking and exam grading. AI-powered systems can analyze student behavior, identify learning patterns, and provide personalized feedback, creating an adaptive learning environment that responds to each student's needs. This study aims to review current research on the use of modern AI technologies such as machine learning, deep learning, and natural language processing in education. The study summarizes the challenges associated with these technologies. Additionally, the study suggests improvements for each of the challenges discussed, focusing on creating more robust, transparent, and equitable AI systems. It also summarizes the key databases used to build AI models in education, highlighting their role in enhancing the accuracy and efficiency of AI-driven educational tools. Despite the vast advancements in AI-based educational technologies, there remains a critical gap in understanding how these systems can equitably benefit diverse student populations and ensure ethical implementation. This study not only examines existing AI applications but also critically evaluates their limitations, particularly in ensuring fairness, reducing biases, and maintaining transparency in decision-making. By addressing these concerns, this research contributes to the ongoing discourse on improving AI-driven education systems to better serve students from various backgrounds and learning capabilities. This study is expected to contribute to the development of AI applications that not only improve the quality of learning but also provide a more efficient, equitable, and interactive learning environment that supports both educators and students.
Keywords:	Artificial Intelligence; Sustainable Education; Deep Learning; Machine Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	AUTOMATED SKIN LESION SEGMENTATION THROUGH HARMONIZING DERMOSCOPIC IMAGES WITH FUSION-BASED WHITE BALANCING
Author:	P. SHOBANA , J. DENY, P. SWAPNA
Abstract:	To resolve aberrations resulting from the dominating single color channel in the RGB plane, this work presents a novel approach for dividing skin lesions in dermoscopic pictures. The accuracy of melanoma diagnosis is greatly impacted by these abnormalities; hence a thorough approach is required. The suggested method combines undesirable hair removal, white balance, and picture segmentation based on unsupervised learning. Two different picture versions are produced by using the Colour Equalisation approach and Shades of Grey Method after starting with the iterative Dull Razor approach for efficient depilation. Dermoscopic image analysis is enhanced by integration using a multi-scale image fusion approach, which promises better benign lesion classification and melanoma detection accuracy. By combining two color-corrected versions, creating conclusive skin lesion delineation by k-means clustering, and leveraging inherent textural information acquired through the Gabor filter, the multi-scale image fusion approach further improves the process. A very detailed picture of the skin lesion is produced by combining the resultant three-segmented images, demonstrating the complexity of this cutting-edge dermoscopic image processing technique.
Keywords:	Dermoscopic images, Skin lesion, Melanoma, Color correction, Image fusion, Color artifacts
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	DESIGN AND ANALYSIS OF SPIN-BASED LOGIC GATES FOR ENHANCING COMPUTATIONAL EFFICIENCY
Author:	PRITI JAGATSING RAJPUT , SHEETAL U. BHANDARI , GUL FAROZ AHMAD MALIK
Abstract:	Spin transistors (Spin-FETs) are a promising way to solve problems with non-volatility, speed, and power consumption in standard CMOS technology. The Spin-FET electrical model using InAs channel material with 800 nm and 820 nm channel lengths is briefly explained in this paper. Spin-FET local geometry can be used to execute digital logic functions according to the experimental analysis. The local geometry of the spin transistor is designed using the Datta-Das concept and the spin injection and detection theory. It is implemented using the Verilog-A language on a Cadence platform. Transient and DC analysis for various channel lengths have been used to validate the results. The power consumption and capacitance of the complementary spin-FET and conventional CMOS have also been compared. This permits the inverter and buffer gate to function with a complimentary spin-transistor. In contrast to other research, our study presents XNOR and XOR gates that are built using a single gate SFET control input technique. However, our design improves computational efficiency by using a novel technique to regulate spin-based logic gates.
Keywords:	Spin-FET Modeling, Local Geometry, Beyond CMOS Technology, Single Gate Spin-FET, Logic Gates, Complementary Spin-FET.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	AN IMPROVED XCEPTION BASED FEATURE LEARNING AND HONEYBADGER OPTIMIZED LONG SHORT-TERM MEMORY FOR ACCURATE BRAIN TUMOR CLASSIFICATION WITH DATASET BALANCING APPROACH
Author:	PARVATHY JYOTHI , DR.S.DHANASEKARAN
Abstract:	A brain tumour (BT) is a severe and lethal disease that significantly reduces human lifespan. Magnetic resonance imaging (MRI) is a commonly employed imaging technique for the early detection of tumours. The segmentation and classification of brain tumours (BTs) through manual and traditional methods is a labour-intensive and subjective process, necessitating the involvement of expert radiologists for evaluation, which may result in prediction inaccuracies. Class imbalance presents a notable challenge in MRI datasets, impacting the efficacy of the classification system. This paper presents an optimised deep learning approach, referred to as Honey Badger optimised Long Short-Term Memory (HBLSTM), aimed at the detection of BT through effective segmentation and feature extraction methodologies. The preprocessing steps are conducted on the collected CE-MRI dataset to reduce noise through Gaussian filtering. The class imbalance issue is addressed through the application of the Adaptive Synthetic Sampling (ADASYN) model. The system employs the Spatial and Channel attention-based Three-Dimensional U-shaped Network (SC3DUNet) for tumour segmentation. The segmentation images utilise the most discriminative features through the Spatial Pyramid Pooling centred Xception Network (SPPXNet). The essential features are subsequently selected utilising the Diagonal Linear Uniform and Tangent Flight-based Butterfly Optimisation Algorithm (DTBOA). The tumour classes are classified utilising the HBLSTM algorithm. The experimental results demonstrate the efficacy of our hybrid deep learning models, achieving an average accuracy of 99.81% in tumour detection, surpassing current state-of-the-art models.
Keywords:	Brain Tumor, Segmentation, Magnetic Resonance Imaging, Figshare Dataset, Dataset Balancing, Pre-trained CNN, Deep Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	ADAPTING LEARNING IN A CONTEXT-AWARE MOBILE LEARNING SYSTEM USING THE DYNAMIC CASE BASED REASONING
Author:	NIHAD EL GHOUCH, AZIZ MAHBOUB
Abstract:	Artificial intelligence has become an increasingly present technology in our daily lives and mobile applications are no exception. Artificial intelligence offers many advantages in the development of mobile applications, on the one hand improving the user experience and on the other hand opening up new possibilities for developers. Learning is one of the areas concerned with the integration of artificial intelligence in the development of mobile applications, it is mobile learning which consists of using mobile devices to access content and educational resources at anytime and anywhere by adapting and creating individualized learning paths based on each learner's needs, interests and prior knowledge. This article presents an approach to an adaptive learning system designed for mobile devices. This approach allows real-time personalized monitoring of the learner during the learning process, providing a personalized learning path based on observation of the context and centered on the learner using mobile devices. This monitoring is carried out using the Felder-Silverman learning style to detect the learning style of each learner, and the artificial intelligence approach of Case-Based Reasoning to ensure automatic prediction and adaptation to the dynamic changes in the behavior of the learner during the learning process (their profile and the characteristics of their mobile device) through the search for similar past learning paths.
Keywords:	Case-Based Reasoning, Mobile Devices, E-Learning, Artificial Intelligence, Adaptive Learning System, Context-Aware, Mobile Learning System, K-Nearest Neighbours
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	A COMPARATIVE ANALYSIS OF CNN, GA, RF & RNN FOR IMAGE CLASSIFICATION: INSIGHTS ON PERFORMANCE AND OPTIMISATION USING HYBRID APPROACHES
Author:	T. ANGAMUTHU , A. S. ARUNACHALAM
Abstract:	Sugarcanе is a glοbally significant cash crοp, cοntributing tο sugar prοductiοn, biοfuеl dеνеlοpmеnt, and νariοus industrial applicatiοns. Hοwеνеr, its prοductiνity is sеνеrеly affеctеd by fungal, bactеrial, and νiral disеasеs, lеading tο substantial еcοnοmic lοssеs. Traditiοnal disеasе idеntificatiοn mеthοds, such as manual fiеld inspеctiοns and biοchеmical analysis, arе οftеn labοr- intеnsiνе, timе-cοnsuming, and prοnе tο human еrrοr. Thе adνеnt οf dееp lеarning has rеνοlutiοnizеd disеasе dеtеctiοn in prеcisiοn agriculturе, but еxisting standalοnе mοdеls facе challеngеs rеlatеd tο cοmputatiοnal еfficiеncy, fеaturе еxtractiοn, and gеnеralizatiοn ability. Tο addrеss thеsе challеngеs, this study prοpοsеs a hybrid dееp lеarning framеwοrk that intеgratеs Cοnνοlutiοnal Nеural Nеtwοrks (CNNs) fοr rοbust fеaturе еxtractiοn, Rеcurrеnt Nеural Nеtwοrks (RNNs) fοr capturing tеmpοral dеpеndеnciеs in disеasе prοgrеssiοn, Gеnеtic Algοrithms (GAs) fοr hypеrparamеtеr οptimizatiοn, and Randοm Fοrеst (RF) fοr еnhancеd classificatiοn pеrfοrmancе. Thе prοpοsеd mοdеl was trainеd and tеstеd οn a datasеt cοnsisting οf 3,750 sugarcanе lеaf imagеs catеgοrizеd intο multiplе disеasе classеs. A randοmizеd stratifiеd split was usеd tο еnsurе balancеd training (70%) and tеsting (30%) data distributiοn. Еxpеrimеntal rеsults indicatе that thе hybrid mοdеl significantly οutpеrfοrms cοnνеntiοnal dееp lеarning classifiеrs. Thе prοpοsеd CNN-GA-RNN-RF hybrid framеwοrk achiеνеd an accuracy οf 92.5%, οutpеrfοrming standalοnе CNN (89.3%), RNN (90.2%), GA-οptimizеd CNN (91.1%), and RF- basеd classifiеrs (87.8%). Thе mοdеl alsο dеmοnstratеd supеriοr prеcisiοn (0.93), rеcall (0.91), and F1-scοrе (0.92), cοnfirming its rοbustnеss in distinguishing bеtwееn hеalthy and disеasеd lеaνеs. Furthеrmοrе, cοnfusiοn matrix analysis rеνеalеd a substantial rеductiοn in falsе pοsitiνеs and falsе nеgatiνеs, еnhancing thе mοdеl’s rеliability fοr rеal-wοrld dеplοymеnt. By cοmbining dееp lеarning with еνοlutiοnary οptimizatiοn and еnsеmblе lеarning, this study prονidеs an AI- driνеn, scalablе, and high-pеrfοrmancе apprοach fοr autοmatеd sugarcanе disеasе dеtеctiοn. Thе findings haνе significant implicatiοns fοr prеcisiοn agriculturе, еnabling farmеrs and agricultural stakеhοldеrs tο dеtеct disеasеs at an еarly stagе, minimizе crοp lοssеs, and οptimizе disеasе managеmеnt stratеgiеs. Futurе rеsеarch will еxplοrе mοdеl gеnеralizatiοn acrοss diνеrsе еnνirοnmеntal cοnditiοns and intеgratiοn with еdgе cοmputing dеνicеs fοr rеal-timе fiеld applicatiοns.
Keywords:	Sugarcanе Disеasе Dеtеctiοn, Hybrid Dееp Lеarning, Cοnνοlutiοnal Nеural Nеtwοrks, Rеcurrеnt Nеural Nеtwοrks, Gеnеtic Algοrithms, Randοm Fοrеst, Prеcisiοn Agriculturе, Smart Farming.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	CLASSIFICATION OF SATELLITE IMAGES USING ADVANCED TOKENS-TO-TOKEN TRANSFORMER WITH PSO OPTIMIZATION
Author:	RONDI PUSHPA LATHA , DR PERSIS VOOLA
Abstract:	Satellite image classification plays a pivotal role in diverse applications, including land use monitoring, urban planning, and environmental analysis. This paper explores the comprehensive classification of satellite images into five classes: desert, forest, green fields, oceans, and urban areas. Initial preprocessing techniques such as resizing, histogram equalization, noise reduction, rotation, cropping, color jittering, and random erasing were applied to enhance data quality. Four Transformer Neural Network (TNN) models i.e., Vision Transformer (ViT), Class Attention Image Transformer (CAiT), Pyramid Vision Transformer (PVT), and Tokens-to-Token Vision Transformer (T2T-ViT) were analyzed. Among these, T2T-ViT demonstrated the best accuracy at 73.21%. Further optimization of T2T-ViT using machine learning techniques, including ensemble methods, feature scaling, and stratified k-fold cross-validation, achieved an accuracy of 84.09%. Subsequently, Particle Swarm Optimization (PSO) was employed for hyperparameter tuning, boosting the model accuracy to 98.75%. This research highlights the efficacy of combining advanced TNN architectures with optimization strategies for robust satellite image classification.
Keywords:	Satellite Image Processing, Vision Transformers, Preprocessing, Soft Computing Techniques.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	CONSUMER BEHAVIOR ON THE INTERNET: METHODOLOGY, TOOLS AND CURRENT TRENDS
Author:	FAWZIEH MOHAMMED MASAD , NASIEM MOHAMMED ALYAMI , HASSAN ALI AL-ABABNEH , OLGA POPOVA , NATALIYA MASLOVA , OLEKSANDR TYSHKO
Abstract:	The key objective of the article is to develop methodological approaches to identifying modern trends in consumer behavior on the Internet. The object of the study is marketing concepts and their transformation under the influence of modern Internet technologies, which are reflected in consumer behavior, which requires study to determine strategic directions and guidelines for marketing. The relevance and necessity of studying this issue is due to the fundamental penetration of technology and the Internet and the impact on consumer behavior, which is a key link in marketing and their transformation should be taken into account in the marketing strategy of modern companies. The main results of the study are characterized by the following: the main types of consumer behavior models are systematized, which made it possible to highlight their features, which will take into account all their specifics when forming a marketing strategy for modern companies; the key stages of developing marketing concepts are structured and the role of consumers in each of them is argued; The trends of consumer behavior on the Internet are conceptualized from the point of view of socio-demographic and gender characteristics, as well as their propensity to make purchases on the Internet. The presented results of the study made it possible to determine modern trends in consumer behavior based on the developed methodology and tools of economic and statistical analysis of daily marketing models, development of its key tools and assessments of the influx into the behavior of the lull. Practical application of the results will ensure the effectiveness of the formation of marketing policy and strategy, taking into account the developed aspects in terms of studying consumer behavior, which is key in modern marketing under the influence of modern Internet technologies and innovations.
Keywords:	Marketing, Strategy, Model, Consumer Behavior
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	CYBERATTACK PREVENTION AND DETECTION IN SMART POWER SYSTEMS USING DEEP LEARNING
Author:	BADDU NAIK B , MANAM RAVINDRA , SIMHADRI MALLIKARJUNA RAO , SRIKANTH KILARU , MADAMANCHI BRAHMAIAH , BEZAWADA MANASA , MURALIDHAR V
Abstract:	Cyber security in power systems is of paramount importance due to the critical nature of these infrastructures. With the increasing digitization of power systems, ensuring cyber security has become imperative to safeguard critical infrastructure. This paper investigates the utilization of meta-heuristic and deep learning algorithms to bolster cyber security in power systems. Traditional supervised machine learning algorithms, including Artificial Neural Networks (ANNs), Convolutional Neural Networks (CNNs), and Support Vector Machines (SVMs), are benchmarked against the proposed algorithm to assess their effectiveness. The proposed algorithm optimizes the hyper parameters and architectures of deep learning models, thereby improving their performance in detecting cyber threats. Cyber-attacks on power systems can have severe consequences, ranging from service disruptions to cascading failures with widespread societal impacts. This research paper investigates the integration of meta-heuristic and deep learning algorithms to enhance cyber security in power systems. Meta-heuristic algorithms offer efficient optimization solutions, while deep learning techniques excel in pattern recognition and anomaly detection. By combining these approaches, a comprehensive framework is proposed for threat detection and mitigation. The paper reviews existing literature, presents methodologies, and discusses potential benefits and challenges. Case studies and experimental results demonstrate the efficacy of the integrated approach in enhancing cyber security in power systems. This research contributes to the advancement of robust and adaptive cyber security measures for critical infrastructure protection.
Keywords:	Cyber Security, Power Systems, Meta-Heuristic Algorithms, Deep Learning, Artificial Neural Networks (ANNs), Convolutional Neural Networks (CNNs), and Support Vector Machines
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	THE IMPACT OF ARTIFICIAL INTELLIGENCE ON THE DEVELOPMENT OF THE DIGITAL BUSINESS ECOSYSTEM
Author:	ARTUR ZHAVORONOK , SVІTLANA FILYPPOVA , YULIIA TOCHYLINA , KATERYNA OZARKO , SERGEY NEYKOV , DENYS KRYLOV
Abstract:	In the article, main aspects of the impact of artificial intelligence on functioning and development of the digital business ecosystem are considered. Urgency of using digital technologies, in particular artificial intelligence, for business development is substantiated. Experience of implementing artificial intelligence in functioning of business ecosystems and their main components in Ukraine is analyzed, and main challenges by using artificial intelligence technologies are identified. Based on the study of the essence of phenomenon of “digital business ecosystem” and effects of implementing artificial intelligence in functioning of digital ecosystems, the model of the digital business ecosystem is proposed, and positive features of the impact of artificial intelligence on development of the digital business ecosystem are highlighted. It is proven that existing problems in using artificial intelligence can be solved based on the strategic approach, which is the key one in development of effective modern policies in digitalization of the economy and management. It is substantiated that effectiveness strategic measures depend on coordination of the mechanism of the impact of artificial intelligence on the business ecosystem development. The model of the specified mechanism is proposed, and features of its action are determined.
Keywords:	Business, Business Ecosystem, Strategic Management, Digitalization, Digital Economy, Digital Technologies, Artificial Intelligence.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	A NOVEL APPROACH FOR BREAST CANCER PREDICTION USING IMPROVED GATED RECURRENT UNIT
Author:	TINTU P B , DR. S VENI
Abstract:	The cells that make up the breast tissue can transform into a tumorous malignancy known as breast cancer. Although males are not immune, it is more frequent in women and is among the most common malignancies globally. In most cases, the illness starts in the breast's ducts or lobules but, if left untreated, can metastasize to other organs. Predicting the likelihood of breast cancer is essential for both early detection and treatment planning. This study presents an advanced approach to breast cancer prediction using an Improved Gated Recurrent Unit (GRU) model. The methodology begins with preprocessing a CSV dataset of numerical records, where Recursive Feature Selection with Extra Tree Classifier (RFET) is employed to identify the most relevant features, enhancing the model's predictive accuracy. Following feature selection, an Improved GRU model is utilized for classification and predictive modelling. The improved GRU architecture incorporates optimizations to improve learning efficiency and accuracy, utilizing temporal dependencies within the data. High predictive performance was achieved by the suggested method, according to the findings, providing a useful tool for early identification and diagnosis of breast cancer.
Keywords:	Breast Cancer, Early Diagnosis, Extra Tree Classifier, Improved Gated Recurrent Unit, Recursive Feature Selection
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	ADDRESSING THE CHALLENGES OF REAL-TIME OBJECT RECOGNITION AND NAVIGATION IN AUTONOMOUS SYSTEMS: A HYBRID SENSOR FUSION APPROACH
Author:	Mr. VOREM KISHORE , Mr. S R V PRASAD REDDY , Mrs. CHALAMALA HARIKA DIVYA , Mr. MULAKA MADHAVA REDDY , Dr. N. SATHEESH , SRINIVASA RAO MADALA , Mr. GARIGIPATI RAMA KRISHNA
Abstract:	The growing demand for efficient and accurate navigation systems in autonomous applications such as drones, self-driving vehicles, and robotics has necessitated the development of advanced detection and recognition devices. This paper proposes a hybrid model combining a Navigation Detection Device (NDD) with an Intelligent Object Recognizer (IOR) to enhance both the accuracy of navigation and object detection. The hybrid model leverages state-of-the-art sensor fusion techniques, machine learning algorithms, and real-time processing to ensure reliable and precise navigation, even in dynamic and unpredictable environments. Our proposed system integrates various sensing modalities, including LiDAR, GPS, and cameras, with deep learning-based object recognition models such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs). This allows the device to not only detect and classify objects but also to predict their movements and adjust the navigation path accordingly. One of the key innovations of the hybrid model is its ability to use sensor fusion to compensate for weaknesses in individual sensors. For instance, the combination of LiDAR and vision systems ensures high accuracy in object recognition, even in low-light conditions, while GPS and inertial measurement units (IMUs) provide robust positional data for precise navigation. The system is designed to operate in real-time, making it suitable for high-speed applications where quick decision-making is critical. Additionally, the hybrid model incorporates a self-learning mechanism, enabling it to adapt to new environments and improve performance over time. Performance evaluations demonstrate significant improvements in both navigation accuracy and object recognition capabilities compared to conventional systems. The proposed hybrid model is poised to be a transformative solution for autonomous systems, offering enhanced safety, reliability, and efficiency.
Keywords:	Navigation Detection, Object Recognition, Sensor Fusion, Autonomous Systems, Machine Learning.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	GRAPH-BASED FRIEND RECOMMENDATIONS ON SOCIAL MEDIA USING MACHINE LEARNING
Author:	SRINIVASA RAO. D , RAMESH BABU.CH , SRAVAN KIRAN.V , M.KOTESWARA RAO , REVATHI. A , RAJASEKHAR. N
Abstract:	Social media platforms facilitate the exchange of vast amounts of information within their networks, where users and their friends form vast communities. To foster meaningful connections, recommendation systems play a vital role in linking users with similar interests. Existing social media networks rely on users' networks to suggest friends, but this approach may not accurately capture users' real-life preferences. This study presents a novel Facebook friend recommendation system utilizing the XGBoost algorithm, which analyzes users' social graphs, incorporates diverse connection parameters, and evaluates various similarity measures. Specifically, we employ similarity measures such as Jaccard distance and the Otsuka-Ochiai coefficient to filter graph information and calculate user ratings. The model leverages features extracted from Adar Index, Hits Score, Katz Centrality, and Page Rank. Our proposed system, utilizing both Random Forest and XGBoost algorithms, achieves 99.99% accuracy. Notably, XGBoost excels in efficiency and speed, making it an ideal choice for large-scale social networks. By suggesting friends with similar ratings, our system enhances user experience and fosters meaningful connections within the Facebook community.
Keywords:	XGBoost, Friend Recommendation System, Adar Index, Hits Score, Katz Centrality, Page Rank, Jaccard Distance, Otsuka-Ochiai Coefficient.
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text

Title:	THE EFFECTIVENESS OF USING MOBILE SOCIAL MEDIA IN LEARNING AND TEACHING ENVIRONMENTAL SCIENTIFIC CONCEPTS IN THE HOLY QURAN
Author:	MAHMOUD ALSALTI, MOHAMMAD ALTARAWNEH, BUSHRA AL-DRAIDI
Abstract:	This research paper pinpoints the effectiveness of using mobile social media in learning and teaching environmental scientific concepts in the Holy Qur'an from Al-Zaytoonah University of Jordan students’ perspective and their attitudes toward it. The quasi-experimental research method is utilized to achieve the research objectives. The research sample consists of 34 male and female students purposefully selected from the Classroom Teacher Department at the Faculty of Arts at Al-Zaytoonah University of Jordan. The research sample members are randomly distributed as follows: An experimental group with (17) female students, and a control group with (17) male students. The research instrument consists of a measure of students’ attitudes toward using the mobile social media method in learning and teaching scientific environmental concepts in the Holy Qur’an. The findings indicate statistically significant differences between the mean of the scores of the experimental group and the scores of the control group in favor of the experimental group, as the F-value is (193.891), which is a statistically significant value at the level of (0.000).
Keywords:	Concepts, Environment, Learning, Mobile, Social Networking
Source:	Journal of Theoretical and Applied Information Technology 15^th May 2025 -- Vol. 103. No. 9-- 2025
Full Text