		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
December 2024 | Vol. 102 No.24

Title:	DEVELOPMENT OF ONTOLOGY FOR RUBRIC ASSESSMENT USING METHONTOLOGY
Author:	NUR FADILA AKMA MAMAT, NOOR MAIZURA MOHAMAD NOOR, ROSMAYATI MOHEMAD, NOOR AZLIZA CHE MAT, DADABAEV SARDORBEK, FOZILJANOVA MARXABO, KANEEKA VIDANAGE
Abstract:	In a variety of fields, including management, medicine, business, education, and others, sophisticated technology that facilitates access to relevant data is essential for supporting decision-making and resolving challenging issues. The absence of explicit representation of knowledge and data modelling through standards like RDF and OWL continues to plague the rubric evaluation sector and hinder the effective sharing between expert and general users. In order to properly characterize important aspects and norms, covering the characteristic of rubric and knowledge of psychomotor skills level, this article presents the ontology known as Psychomotor Learning Domain (PLD) ontology. Creating and developing an ontology model for rubric assessment is the goal. There are several ways to construct ontologies. A well-defined and organized methodology can shorten the time needed to construct an ontology, increasing the likelihood that the project will succeed. METHONTOLOGY is used in the specification, conceptualization, formalization, implementation, and maintenance phases of the ontology building process. Adherence to relevant norms and laws has been widely established during the ontology creation. Only the conceptualization—the process of organizing knowledge for ontology implementation—is the subject of this paper. The conceptualization of the tasks in the task set for knowledge structuring is included in methodology. It makes ontologies buildable at the knowledge level. The developed ontology is meant to serve as a domain knowledge base for further programs, including expert systems.
Keywords:	Rubric Assessment, Ontology, Higher Education, Psychomotor Learning Domain
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	MAKING THE VISUALLY IMPAIRED FEEL IMAGES USING DLEVIS (DEEP LEARNING EYE FOR VISUALLY IMPAIRMENTS)
Author:	S. MURUGESAN, DR. N. BALAJIRAJA
Abstract:	In the world, 25% of the 285 million persons who suffer from visual impairments (VIs) are totally blind. The thrill of sight is also taken away from people with VIs. Even through visuals, they are unable to appreciate nature, green spaces, or surroundings. Furthermore, social media and recent technological advancements have served as the primary means of bringing people together on a worldwide scale. Common people's lives are impacted by mobile social media, and their usage patterns are the subject of in-depth research. Users with VIs as special user groups, however, are frequently disregarded. Few studies have been done to find out how they connect with popular mobile social media today. By proposing a method that interprets images and provides information to individuals with VIs, this study aims to close these gaps and enable their introduction as a unique user group on social media. The suggested method, which is based on DLTs, is dubbed DLEVIs (Deep Learning Eye for Visually Impairments). The accuracy of the scheme is also assessed in the study, where the findings show relatively low categorization mistakes.
Keywords:	Deep Learning, Visually Impaired, Mobile Social media, Captioning Images, Social Interactions, Social Networking Sites
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	THE USE OF BIG DATA IN THE DETECTION OF ECONOMIC CRIMES IN PUBLIC PROCUREMENTS
Author:	HASSAN ALI AL-ABABNEH, YAROSLAV FEDORCHUK, ANDRIY TYMCHYSHYN, GENNADY PISHCHENKO, SERHII HRYTSAI
Abstract:	The article deals with the role of digital transparency tools in the fight against corruption, especially in public administration. The research aimed to assess how such technologies as blockchain, open government data, e-procurement, and e-government services increase transparency and accountability while reducing corruption risks. The study fills a gap in the literature by proposing a new approach to combating corruption through digital solutions. The study covers 10,000 procurement contracts from the European Union (EU), using modern methods such as Big Data (BD) analytics, statistical analysis, simulation model. These approaches have significantly improved the detection of fraud-related anomalies compared to traditional methods based on manual checks and limited data. The study compares the effectiveness of digital transparency initiatives in several countries, revealing the conditions under which these tools are most effective. Besides, such challenges as insufficient digital literacy and barriers to implementation are discussed. The results showed that digital transparency helps to reduce corruption by improving access to information and simplifying processes, but its success depends on political and economic infrastructure. The article provides recommendations for policymakers on the effective use of digital tools in different governance systems, significantly contributing to understanding the interaction of technology and anti-corruption reforms.
Keywords:	Economic Crimes, Public Procurement, Fraud, BIG DATA, Network Analysis.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	IMPROVED POWER OPTIMIZATION TECHNIQUES IN MOBILE AD HOC NETWORK
Author:	S. HEMALATHA, TAVANAM VENKATA RAO, G.S. N MURTHY, KHADRI SYED FAIZZ AHMAD, P. SUPRIYA, DR. CHARANJEET SINGH
Abstract:	Mobile Ad hoc Networks (MANETs) are dynamic, self-organizing networks vital for communication in scenarios without established infrastructure, such as disaster recovery and military operations. However, MANETs face critical challenges, including power management, collision issues, and routing inefficiencies caused by hidden and exposed terminal problems. This study proposes an integrated solution combining Beam Sector Directional Antennas with a Mutual Exclusion Medium Access Control (ME-MAC) protocol. The proposed system optimizes power consumption and improves network performance through precise beam-based packet transmission and collision avoidance mechanisms. Simulations conducted using NS2.32 reveal significant improvements over traditional Omni-Directional Antennas. Specifically, signal power consumption is reduced by 50%, collision rates by 25%, energy consumption by 30%, and end-to-end delay by 45-50%. Furthermore, the PMBS-Antenna achieves 15% higher transmission speed, 10% better efficiency, and a 40-50% increase in packet delivery ratio. These findings demonstrate the potential of the proposed approach to address longstanding MANET challenges, paving the way for more efficient and reliable communication systems.
Keywords:	MANET,Antenna, Physical Layer, MAC layer, Hidden and Exposed Terminal (HET) problem, ME-MAC, protocol
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	ENHANCING CLARITY IN CHINESE SOFTWARE REQUIREMENTS: A BOILERPLATE-BASED APPROACH TO IMPROVE REQUIREMENT EXPRESSION
Author:	HE JIAYING, MOHD HAFEEZ OSMAN, SA'ADAH HASSAN, NG KENG YAP
Abstract:	During the rapid process of software development, requirements engineering plays a crucial role in ensuring the success of software projects. Clear and accurate requirements specification directly impact the effectiveness of project implementation, influences customer satisfaction, and ultimately determines the overall success rate of the project. However, ambiguity and structural issues are common in Chinese software requirements, which can hinder understanding and implementation. These issues not only increase the complexity of project execution but may also lead to extended development cycles and wasted resources. To address these challenges, this paper aims to develop a set of Chinese-based software requirement boilerplate to improve the clarity and accuracy of requirements specification. The boilerplate is designed with consideration for Chinese grammar conventions and international standards, offering a universal and user-friendly solution. This paper reviews related literature, discuss the limitations of current requirements specification method, and details the boilerplate design process and evaluation methodology. The effectiveness of the boilerplate is demonstrated through the reconstruction of requirements statements across various software industries, complemented by expert assessments and feedback from industry professionals. This study provides a practical foundation for Chinese software requirements specification, aiming to enhance the quality and efficiency of requirements expression.
Keywords:	Software Requirements Engineering, Chinese Requirements Expression, boilerplate Design, Clarity, Requirements Normalization
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	DEEP LEARNING-BASED CONGESTION CONTROL IN VLSI FOR PLACEMENT AND ROUTING
Author:	SHAIK ASIF HUSSAIN, SHAIK KARIMULLAH, FAHIMUDDIN SHAIK, SYED JAVEED BASHA
Abstract:	This work presents the application of a recurrent neural network based on the deep learning technique to evaluate design constraints for routing and placement flow, with the assistance of existed floorplan approach, to estimate the value of congestion. Effective area utilization is essential in very large-scale integration (VLSI) circuit design, wherein congestion reduction is also associated with improving floorplan, placement, and routing. This improvement significantly helps a circuit’s compact design and performance. Congestion is a fundamental key issue in VLI for estimating the density of area underlies routing among various computational blocks. We used the deep learning technique for the simulation of the standard benchmark circuit’s planned area for better placement. Prior approaches estimated the values of congestion for standard architectures, whereas this work considered the floorplan and placement outputs of standard MCNCBM Circuit using the recurrent neural network-based algorithm (RNN-based algorithm), which yielded better results for placement and routing of VLSI circuits and simulated it to estimate the congestion for the circuit design. A recurrent neural network, an artificial neural network, uses sequential data or time series data. To handle problems that involve order or time, such as image captioning, language translation, and speech recognition, deep learning algorithms are frequently utilized. Learning occurs through the training data in RNNs, just as it does in feed-forward RNNs and convolutional neural networks (CNNs). The information "remembered" from earlier inputs can then be used to modify the current input and the output. This sets them apart from other systems. As they are currently understood, deep neural networks work under the assumption that inputs and outputs are independent of one another. On the other hand, the output of recurrent neural networks is, nevertheless, influenced by the information processed in the network before it. Although the occurrence of events in the future might help in determining the outcome of a particular sequence, unidirectional recurrent neural networks cannot consider these occurrences when generating predictions.
Keywords:	Congestion, Placement, RNN-based algorithm, VLSI, Partial blockage technique.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	EMPIRICAL SOUNDINGS TO FACIAL COUNTENANCE RECOGNITION USING CNN
Author:	DR.S. SAI KUMAR, DR.SURYA PRASADA RAO BORRA, DR.YASWANTH KUMAR ALAPATI, P UDAYA BHANU, DR.K. RAMESH CHANDRA, BAKKALA SANTHA KUMAR
Abstract:	There are many uses for the devices, and the number of uses is growing every day. Machine perception will assist in carrying out a variety of activities, including intricate ones. Machine perception enables machines to comprehend both their physical surroundings and the conversation partner's intentions. In this study, we classified the images into categories such as happiness, sadness, anger, amazement, dislike, and anxiety using deep learning techniques such convolution neural networks. This method is used because CNN produces better results than other statistical techniques. Using CNN requires feature learning, which is a crucial task. Additionally, the community was assessed using two corpora: one was used for community education and the other was used for defining the structure of a network. The network that produced results with first-class accuracy was compared to the second dataset. When tested on a unique fact set that shows facial emotions, the network mentioned favourable outcomes even though it had been trained using the best corpus. Although the results showed that the network was no longer king, the evidence shows that deep learning is likely appropriate for categorizing facial emotion expressions. As a result, deep mastery can enhance human-system connection since its learning skills will enable machines to perceive more.
Keywords:	CNN, Open CV, Facial Expression.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	LSTM ADAPTIVE HYPERPARAMETER TUNING FOR FINANCIAL TIME SERIES FORECASTING USING CUSTOM GRADIENT-BASED METHODS: A COMPARATIVE STUDY WITH BAYESIAN OPTIMIZATION
Author:	ADNANE EL OUARDI, BRAHIM ER-RAHA, KHALID TATANE
Abstract:	This paper presents an adaptive hyperparameter tuning system for Long Short-Term Memory (LSTM) models, focusing on a gradient-based approach to achieve efficient and precise optimization. The proposed system is designed to dynamically adjust critical hyperparameters such as the learning rate, number of units, dropout rate, and batch size during the training process. By leveraging gradient information, the system iteratively refines the hyperparameter values, aiming to minimize the Mean Squared Error (MSE) and enhance the model's predictive accuracy. To evaluate the effectiveness of the gradient-based approach, the study includes Bayesian optimization as a benchmark, a method widely recognized for its probabilistic framework and established success in hyperparameter tuning. Comparative analysis is conducted in terms of loss metrics, including MSE and additional performance indicators, as well as execution time, highlighting the trade-offs between optimization efficiency and computational cost. The results demonstrate the gradient-based system's ability to adapt to high-dimensional and complex hyperparameter spaces with reduced computational overhead, while consistently achieving superior performance. The experiment is applied to the challenging task of forecasting the S&P 500 index, a real-world financial time series problem that demands robustness and precision.
Keywords:	Hyperparameter Optimization, Gradient-Based Tuning, LSTM Networks, Bayesian Optimization, Time Series Forecasting
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	CLASSIFICATION OF CORN LEAF DISEASES USING CNN: A DEEP LEARNING APPROACH
Author:	HANDRIZAL, FUZY YUSTIKA MANIK, VICTORY J SIANTURI
Abstract:	One of the major problems that needs to be addressed is the identification of infection in corn leaves. The quality and production of crops can decrease due to the presence of diseases on corn leaves. The process of diagnosing diseases on corn leaves manually takes time and effort. Therefore, a more effective and accurate technique is needed to know the existence of diseases on corn leaves. In this study, a Convolution Neural Network (CNN) is used as a technique for identifying diseases on corn leaves. A machine learning model called CNN can be used to identify characteristics in images. To train CNN, a dataset of corn leaf images labeled with the names of predetermined corn leaf diseases is used. The accuracy rate that has been achieved is 95%, with a precision of around 95%. The recall rate also reaches 95%, while the F1 score reaches 95%. This method of identifying diseases on corn leaves using CNN can be used to improve the efficiency and accuracy of disease identification on corn leaves. This method can also be used to develop an early warning system for diseases on corn leaves.
Keywords:	Disease Identification, Corn Leaf, Convolution Neural Network, Accuracy
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	AN ENERGY AND AREA EFFICIENT IOT ARCHITECTURE FOR BIO-MEDICAL APPLICATIONS
Author:	P V GOPIKUMAR, C RAVI SHANKAR REDDY, R MANIKANDAN, T V HYMALAKSHMI, G VENKATESWARLU
Abstract:	Heart Stroke, cardiac arrest are more prominent deceases assassinating the harmony of the people. Meticulous detection of the abnormalities and myocardial infarctions leads to abrupt change in the functionality of the heart. QRS Complex amplitude and R wave amplitude are significant to detect the abnormalities and sudden cardiac arrest. This paper focused to develop an architecture to meet the challenges of IOT enabled wearable devices. Absolute value curve length transform (A-CLT) is implemented to detect the QRS complex Changes. The proposed methodology nullifies the multipliers and performs well with adders, shifters and comparators. Eventually, the packing density (area) is minimized. This will improve the processing time and minimizes the dissipating power. This paper addressed the complexity in early detection of the strokes and cardiac arrests by analyzing the QRS complex of the ECG signal. Base line drift, high frequency interference (artifacts) are impacting on the signal generation. The proposed methodology curtails those artifacts and improves the performance of the signal detection and interpretation. The area is miniaturized with the A-CLT approach. 93.36 percent of the area is reduced with A-CLT. 77.61 percent of power is minimized with A-CLT methodology. The computation delay is reduced to 79.64 percent. This paper also addressed the sensitivity and predictivity of the QRS complex amplitude meticulously. 99.46 percent predictivity and 99.24 percent sensitivity is achieved with the proposed methodology. The achieved results are validated with the physician and specifically, achieved lossless compression for enhancing the derivative of ECG Signal and entropy encoding. It is observed that the compressed fraction is 2.05 and is validated with MIT-BIH database. The proposed methodology is surpassing the existing methods. The achieved results proved that this A-CLT applied architecture is best fit for wearable devices to prevent abrupt changes in cardiac functionality and to safeguard human from sudden cardiac arrests.
Keywords:	Absolute Value Curve Length Transform (A-CLT), QRS Complex, Quadratic Spline Wavelet Transform, Electrocardiography (ECG)
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	PROPOSAL OF ENHANCING WATER SAFETY- AN AUTONOMOUS ROBOT FOR DROWNING PREVENTION
Author:	S. HEMALATHA, DR. KIRAN MAYEE ADAVALA, S.N. CHANDRA SHEKHAR, PULLELA SVVSR KUMAR, A.R. VENKATARAMANAN, D NAGA MALLESWARI
Abstract:	This paper presents the development of an advanced image recognition robot, dubbed the "drowning robot," designed to identify individuals at risk of drowning using state-of-the-art hardware and software. The primary aim of the project is to enhance water safety. The system leverages powerful computing platforms, such as the NVIDIA Jetson and Intel NUC, paired with high-resolution cameras and specialized sensors to capture and process real-time video data. By employing image processing technologies like OpenCV and deep learning models such as YOLO (You Only Look Once), the robot can detect human figures and unusual movements in aquatic environments. Operating autonomously, the robot offers a reliable solution for emergency response scenarios and can connect to cloud services for further verification. Key performance metrics, including FLOPS, latency, and frames per second (FPS), are assessed to ensure optimal processing speed for quick detection and action. This cutting-edge technology represents a major advancement in both robotics and safety engineering, with the potential to significantly improve rescue operations during drowning incidents and provide timely alerts to enhance public safety.
Keywords:	Drowning Detection, Robotics, YOLO, OpenCV, Object Detection, Surveillance Technology
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	FACTORS INFLUENCING CUSTOMER PURCHASE DECISIONS DURING LIVE STREAMING SHOPPING
Author:	MOH THAHA RIZIEQ HENTIHU, RIYANTO JAYADI
Abstract:	The purpose of this study is to identify the factors influencing customer purchase decisions during live streaming shopping using a quantitative research method. Data was collected through a questionnaire distributed to 441 respondents, of which 433 had previously shopped via live streaming. The results indicated that 2 hypotheses were rejected while 10 were accepted. The rejection of hypothesis H2 indicates that Interactivity does not have a significant impact on Trust, and the rejection of hypothesis H4 shows that Visualization also does not significantly impact Trust. Interactivity, Visualization, Professionalization, System Quality, Information Quality, and Service Quality influence the Social Presence of Live Streaming Shopping. Professionalization impacts Trust. Both Social Presence of Live Streaming Shopping and Trust affect Purchase Decision, which in turn influences Purchase Intention.
Keywords:	Live Streaming Shopping, Social Commerce, Purchase Decision, Consumer Behavior, E-Commerce
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	CAN AI-DRIVEN FEATURES ALONE FOSTER CUSTOMERS’ TRUST IN INDONESIAN AND MALAYSIAN BANKS?
Author:	ROCHANIA AYU YUNANDA, TOTO RUSMANTO, MOHAMMAD ALI TAREQ, NURIL KUSUMAWARDANI SOEPRAPTO PUTRI
Abstract:	Artificial intelligence (AI) has rapidly transformed numerous industries, including banking. Despite global economic challenges, the banking sectors in Indonesia and Malaysia have shown resilience and growth, actively adopting digital technologies. This study examines the effect of AI features on trust in banking industry of both countries. Besides AI features, this study explores other predictive factors such as income, gender, experience in using banking services and also frequency in dealing with banking services. Questionnaires were distributed to banking customers in Indonesia and Malaysia. Multiple regression analysis were performed using SmartPLS 4. The findings reveal that AI feature alone do not affect customer’ trust. Some demography factors also play the important roles. The implications of this study are significant for both the banking sector and policymakers. By highlighting the role of AI in enhancing customer trust, the study underscores the importance of integrating advanced technologies to improve operational efficiency and customer experience. Prior studies focus on examining the determinants of technology adoption. This study is one among a very limited number of studies scrutinizing the adopted technological features on the customers’ trust in banking industry.
Keywords:	AI Features, Banking, Trust, Income, Banking Experience, Frequency, Indonesia, Malaysia
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	DYNAMIC GRID-BASED CLUSTERING FOR NON-STATIONARY SPATIO-TEMPORAL EVENT PREDICTION
Author:	M.VASAVI, A.MURUGAN, ,K.VENKATESH SHARMA
Abstract:	Spatio-temporal data is increasingly collected in domains such as urban planning, traffic analysis, and environmental monitoring, in which events change quickly over time as well as across different regions. Traditional spatio-temporal clustering models cannot handle the characteristics of non-stationary processes with quick and unpredictable changes. To overcome these limitations, this study presents a dynamic grid-based clustering method that adjusts its spatial and temporal parameters on the fly to improve the accuracy of prediction and computational efficiency for non-stationary spatio-temporal event analysis. We iteratively update grid sizes and time intervals by considering the density of events and movement patterns of objects to capture easily unseen clusters over time with minimum computing cost. The performance of the proposed method was evaluated on two real-world datasets (i.e., urban traffic data and environmental monitoring data) and compared with not only several modified versions of baseline models like DBSCAN or Spatio-Temporal k-Nearest Neighbor (STKNN) but also a naïve approach based on grid concept. The results showed that our proposed dynamic grid-based model outperformed other approaches in terms of both clustering quality (with Silhouette Coefficients value 0.82) and computational time (1.8 seconds), but had comparable error-rate prediction results (mean square error is 0.015). These achievements confirm that our method can adapt to changes in real-time processing environments to meet the needs for continuous event predictions.
Keywords:	Dynamic Grid-Based Clustering, Spatio-Temporal Data, Non-Stationary Events, Real-Time Prediction, Traffic Monitoring, Environmental Monitoring.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	KNOWLEDGE DISCOVERY IN DATABASES FOR HOTEL SERVICE QUALITY IMPROVEMENT THROUGH DATA-MINING APPROACH
Author:	YERIK AFRIANTO SINGGALEN, SIH YULIANA WAHYUNINGTYAS, YOHANES EKO WIDODO, MUHAMAD NUR AGUS DASRA, RUBEN WILLIAM SETIAWAN
Abstract:	This study integrates knowledge discovery in databases (KDD) with sentiment analysis techniques to evaluate customer feedback and improve service quality in the hotel industry. The research emerges from the growing demand for data-driven strategies in the highly competitive hospitality sector, where understanding customer sentiment is crucial for enhancing guest satisfaction and retaining market share. Machine learning models, including Support Vector Machine (SVM), Decision Tree (DT), Naive Bayes Classifier (NBC), and k-nearest Neighbors (k-NN), were employed to extract insights from unstructured text data, identifying key factors that influence guest satisfaction. Results indicated that the SVM model achieved the highest accuracy of 95.4%, with a precision of 93.22% and recall of 95.4%, showcasing its robustness in sentiment classification. In contrast, NBC showed lower effectiveness with 79.09% accuracy, while k-NN faced challenges in complex data scenarios, achieving 60.71% accuracy and an F-measure of 49.99%. The findings suggest that integrating sentiment analysis into hotel management practices can boost customer satisfaction by 20-30%, particularly in staff interaction and facility maintenance. This research presents a novel approach by combining KDD methodologies with qualitative sentiment analysis, moving beyond traditional quantitative metrics to provide deeper insights into customer experiences. However, reliance on online reviews as the primary data source may introduce biases, potentially affecting the generalizability of results. Future research should expand data collection to include structured surveys and direct interviews, incorporate deeper semantic analysis, and utilize real-time sentiment monitoring to enhance service management strategies. This approach could lead to sustainable competitive advantage and continuous improvement in the hospitality industry.
Keywords:	KDD, Hotel, Service Quality, Data Mining
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	APPLYING BLOCKCHAIN TO ENHANCE IN-GAME ECONOMIES IN MULTIPLAYER FPS GAMES
Author:	ANATHA PINDHIKA HERMAWAN, MARIA SERAPHINA ASTRIANI
Abstract:	The contribution of blockchain technology, in particular the Ethereum platform, for enhancing the security, transparency, and efficiency of the transactions happening in multiplayer FPS games, has been examined. In this paper, a blockchain-driven real-time transaction recording system is proposed and implemented based on the Ethereum PoS consensus algorithm to minimize fraudulent activities and data tampering. Extensive testing demonstrated that the integration of blockchain significantly strengthened security measures and transaction accuracy, while maintaining high game performance and seamless player experience. By addressing two of the most common issues found in online gaming-cheating and vulnerability to centralized data-this study shows how blockchain can truly change the face of game economies. The results highlight the feasibility of integrating blockchain technology to provide a decentralized and secure transaction framework, which will further help in creating fairer and more transparent gaming contexts. Further studies should address the scalability of blockchain marketplaces and its impact on player engagement and satisfaction, adding to the growing debate on the importance of blockchain in game development.
Keywords:	Blockchain, Game, First Person Shooter, Ethereum, Game Development,
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	COGNITIVE MODELS OF PATENT TRANSLATION
Author:	LADA BERDNYK, TETIANA VYSOTSKA, SVITLANA KOROTKOVA, NATALIA MOSKALENKO, OLESIA CHERKASHCHENKO
Abstract:	The aim is to assess the effectiveness of cognitive models of translation for the accuracy, speed, and quality of patent translation. The research employed such methods as cognitive modelling, error analysis and the method of contrast analysis. Traditional metrics of translation quality, such as Bilingual Evaluation Understudy and Recall-Oriented Understudy for Gisting Evaluation, were used in the study. The chi-squared test was also used. The results of the study showed that the GPT-3 model with BLEU 0.85 and ROUGE 0.88 showed the best quality indicators of patent translation, providing the highest accuracy and smoothness of translation. The BERT model also performed well with BLEU 0.82 and ROUGE 0.85, preserving the structure and semantics of the original. In contrast, the LSTM and GRU models had lower values — BLEU 0.65 and 0.68, respectively, indicating difficulties with the accuracy of translating specific terms. The study revealed that the GPT-3 model provides the highest translation accuracy of patent texts, followed by BERT. LSTM and GRU models showed medium results, indicating the need for their further optimization. The results confirm the importance of choosing the appropriate model for specific translation tasks. Further research may address issues of improving the quality of text translation using AI tools.
Keywords:	Neural Networks, Chatgpt, Google Gemini, Translation Quality, Error Analysis, Patent Texts.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	APPLICATION OF THE PIECES FRAMEWORK METHOD IN THE ANALYSIS OF USER SATISFACTION LEVELS OF OASIS APPLICATION SERVICES
Author:	IKE KUSDYAH RACHMAWATI, THERESIA PRADIANI, AGUS RAHMAN ALAMSYAH, ABD HADI, SYARIF HIDAYATULLAH
Abstract:	This study aims to analyze the level of user satisfaction of the OASIS (Online School Information System Administration) application service using the Pieces Framework method. This method includes six dimensions: Performance, Information, Economics, Control, Efficiency, and Service, which allows for a thorough evaluation of information systems (Hawkins, 2020). A quantitative approach and survey design were used to collect data from 40 respondents consisting of students, teachers, and education staff in Malang Regency. The specially designed questionnaire uses the Likert scale to measure satisfaction levels. The results of the analysis show that all domains in the Pieces Framework achieved an average score above 4.38, with the Service domain recording the highest score of 4.56. The average total user satisfaction reached 4.43, which is included in the category of "VERY SATISFIED." Although most respondents are satisfied, there is negative feedback that can be used to improve the quality of OASIS app services. These findings provide valuable insights for developers to improve the user experience in the future.
Keywords:	Satisfaction Level, Pieces Framework, OASIS Application, School
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	ACRITICAL REVIEW OF SOFTWARE QUALITY MODELS IN ACADEMIC INFORMATION SYSTEM
Author:	ASMAA JAMEEL AL-NAWAISEH, FADI BANI AHMAD, SABAH JAMIL AL-NAWAISEH
Abstract:	In the contemporary educational landscape, the quality of Academic Information Systems (AIS) like E_learning System plays a pivotal role in enhancing the efficiency and effectiveness of educational institutions. This study examines the critical factors that influence the quality of AIS, including system usability, data accuracy, user satisfaction, and the impact on academic performance. Through a comprehensive analysis of existing literature and empirical data, the research highlights the significance of robust AIS in supporting administrative functions, academic processes, and decision-making. The findings underscore the necessity for continuous evaluation and improvement of these systems to ensure they meet the evolving needs of students, faculty, and administrators. Furthermore, the study provides practical recommendations for educational institutions to optimize their AIS, thereby fostering an environment conducive to academic excellence and innovation.
Keywords:	AIS, Software quality, E_learning, Models, Information System, ISO 9126, CMMI
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	ASSESSING GAMIFICATION GAP IN SPORT APPLICATION FOR INCREASED USER MOTIVATION: AN OCTALYSIS MODEL APPROACH
Author:	WILLIAM YEO , TANTY OKTAVIA
Abstract:	This study investigates the rapid decline in physical activities among Indonesian populations and explores how such a trend contrasts with the overall growth of electronic sports that utilizes gamification to attract user participation and commitment over an extended period. There is numerous gamified sport applications created to increase exercise participation; however, the continued decline is representative of deficiencies in those applications' motivational frameworks. This work, using the Octalysis model, delivers an in-depth analysis of user motivation into eight core gamification drives, based on the results of a survey presented to 101 users of sports applications. The core drives related to intrinsic motivation—in particular, accomplishment and empowerment—present high effectiveness; however, there are still important deficiencies in social engagement, creative expression, and unpredictability elements. These points have serious implications for an urgent need for targeted refinements in gamification strategies to deliver a more engaging, adaptive, socially interactive experience that is better at sustaining user motivation. Indeed, such strategic adjustments have the potential not only to improve retention rates and further enhance user experience but also to contribute toward broader public health challenges through the fostering of sustainable physical activity patterns. This study gives some very important recommendations for app developers and policymakers seeking innovative solutions that are digitally driven to combat sedentary lifestyle practices, therefore helping to advance health and wellness objectives for a diverse user population across Indonesia.
Keywords:	Sport Application, User Motivation, Gap Analysis, Gamification, Octalysis Framework
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	DETECTION OF ABNORMAL FEATURES IN TRANSACTION DATA FOR MALICIOUS ATTACKS USING HIERARCHICAL NETWORK FEATURE EXTRACTION
Author:	DAMODHARAN KUTTIYAPPAN, DR RAJASEKAR V
Abstract:	The growing number and complexity of financial transactions have made the detection of fraudulent activity and cyberattacks a considerable issue for organizations. Conventional rule-based systems and statistical techniques frequently fail to detect complex assaults that masquerade as ordinary transactions. This work presents a deep learning methodology for detecting anomalous aspects in transaction data to efficiently identify potential threats. Feature extraction technique has been used to find the suitable features for further processing the dataset. Deep learning techniques are used to classify transactions as normal or anomalous. This study introduces a deep learning methodology for detecting anomalous transaction data, using advanced techniques like CNNs, RNNs, and Autoencoders to identify potential threats. The model outperforms conventional detection techniques in precision, recall, and F1-score, providing insights into abnormal behavior, and aiding in attack discovery and mitigation. The system evolves through ongoing learning, enhancing its detection precision for changing assault patterns.
Keywords:	Deep Learning, CNN, RNN, Anomalies, Threats, Feature Extraction, Financial Transaction
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	VIRTUAL REALITY AND INTERACTIVE TECHNOLOGIES IN CONTEMPORARY ART: AN ANALYSIS OF CREATIVE OPPORTUNITIES
Author:	NATALIIA YUHAN, LARYSA KORNYTSKA2, DENYS SUCHKOV, ZOYA ALFOROVA, VALERIIA BOIKO
Abstract:	This article aims to explore the transformative impact of virtual reality (VR), artificial intelligence (AI), and interactive technologies on contemporary art. This study provides a new perspective on how these technologies redefine art as an immersive, responsive experience tailored to individual viewer interactions. This research is conducted through a comprehensive literature review of existing studies, articles, and documented cases in which VR, AI, and interactive technologies play a role in artistic practices. Critical studies analyzed include those emphasizing AI's capacity for generating personalized content and VR's ability to create fully immersive environments. The originality of this article lies in its combined focus on VR, AI, and interactivity in contemporary art. By synthesizing these three areas, the article presents an innovative theoretical framework for understanding immersive experiences and contributes to rethinking the viewer's role as an active participant. The findings indicate that these technologies are reshaping art by enhancing engagement and redefining the creative process, offering new modes of collaboration between artists and technologists. Ultimately, this research expands our understanding of integrating modern technology in art, suggesting new approaches to artistic experience and the shifting dynamics between artist, artwork, and viewer.
Keywords:	NFT art, Digital art, Interactive installations, Virtual expositions, Creative technologies, Artistic innovations.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	METHODICAL REVIEW OF MUTATION TESTING FOR SOFTWARE PROJECTS
Author:	MADHAVI KATAMANENI, DR C.S.S. ANUPAMA, CHETLA CHANDRA MOHAN B. BALAJI BHANU, P SWETHA NAGASRI, DR.J MANO RANJINI, PARUCHURI RAMYA, M BALA CHENNAIAH
Abstract:	A fault-based testing technique that has been extensively researched for more than thirty years is mutation testing. Change testing is a rigorous, complex, and expensive testing approach. This testing method intentionally injects incorrect lines of code to test programming's ability to produce results that differ somewhat from the correct or original code. It is a method that ensures the quality of test information by examining if it can identify a set of replacement projects by addressing specific types of defects in the program being tested. Since change investigation is widely regarded as an excellent testing strategy, it is commonly employed to evaluate the test criteria in terms of its transformation ampleness score. The writing on Mutation Testing has contributed an arrangement of methodologies, instruments, improvements, and exact outcomes. This paper gives a complete examination and review of change testing. This investigation gives confirm that Mutation Testing strategies and apparatuses are achieving a condition of development and appropriateness, while the point of Mutation Testing itself is the subject of expanding premium.
Keywords:	Mutation Testing, Mutant, Mutant Adequacy Score, Syntax Errors, Cost, Mutant Operators
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	NEMAEP: A NOVEL ENSEMBLE MACHINE LEARNING FRAMEWORK FOR ACCURATE EFFORT ESTIMATION IN SOFTWARE PROJECTS
Author:	PRATEEK SRIVASTAVA, NIDHI SRIVASTAVA, RASHI AGARWAL, PAWAN SINGH
Abstract:	The increasing complexity of software engineering projects has made accurate effort estimation a formidable challenge. We introduce the NEMAEP framework to address existing methods shortcomings and enhance precision. This framework integrates ensemble learning techniques, specifically the CatBoost gradient boosting algorithm, with Grid Search cross-validation for optimizing hyperparameters. Our robust predictive methodology substantially improves the accuracy and reliability of software effort estimation. We evaluated NEMAEP against well-established regression techniques, including support vector regressor, decision tree, random forest, and multi-layer perceptron. Two datasets were utilized: Cocomo81, comprising 63 projects, and a more extensive China dataset, containing 499 projects. We assessed performance using MAE, RMSE, and R². The NEMAEP methodology demonstrated superior predictive capabilities, achieving accuracy rates of 97% for Cocomo81 and 99% for the China dataset. When applied to the China dataset, the model produced MAE and RMSE scores of 0.0154 and 0.0213, respectively. In Cocomo81, the values were 0.013 and 0.0641. This innovative strategy allows for the effective maximization of resource utilization in software project management. This is achieved by providing project managers with a data-driven tool to navigate modern software development projects.
Keywords:	Software Project Management, Software Effort Estimation, Machine Learning, Ensemble Learning, Grid Search
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	EVALUATING THE IMPACT OF 3D PRINTERS FOR CREATING INNOVATIVE POSTER DESIGNS
Author:	YEVHEN GULA, OKSANA MAZNICHENKO, SVIATOSLAV PODLEVSKYI, ALLA OSADCHA, YELYZAVETA DEREVIANKO
Abstract:	The aim of the study is to substantiate the optimal choice of a software for prototyping and UI/UX design. The aim of the study is to determine the features of using 3D printing, while focusing on the creation of innovative poster designs. The research methods at the initial stage of the study were methods of comparison, calculations of the correlation coefficient, analysis of variance, t-test, calculating Pearson coefficient. It was determined that innovative approaches to creating poster designs are abstract shapes (t=1.783), the use of elements created by artificial intelligence (t=1.772). Illusionary design elements (t=1.743), combinations of mismatched parts (t=1.756), and minimalism (t=1.732) can also be taken into consideration. It was found that the 3D printer from Mcor Technologies has the greatest efficiency in CD printing. The process is related to the realism of the created drawing (93%), the quality of the color scheme (95%), and the scope of application (93%). With the help of the Anycubic Mega Pro printer, it is possible to convey the quality of the composition structure (92%) while providing multi-color printing. It was established that when developing a layout, for high-quality printing it is imperative to take into account color (t=1.783), design elements’ dimensions (t=1.770) as well as the contour complexity (t=1.765). It was found that the obtained findings have a positive practical impact on prospective designers, which is due to the achievement of high results by students of Group 1 in the creation of design projects for 3D printing (92 points). The research objectives are focused on comparing the excellence of design projects when employing interactive versus traditional methods in their creation.
Keywords:	Interactive Technologies, Design Project, Three-Dimensional Elements, Abstract Shapes, Color Scheme.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	REINFORCING QUERY-BASED SURVEILLANCE SYNOPSIS SECURITY THROUGH DISTRIBUTED COMPUTING
Author:	HASSAN I. SAYED AHMED, RASHA SHOITAN, GHADA F. ELKABBANY, MONA M. MOUSSA, MOHAMED S. ABDALLAH, YOUNG-IM CHO
Abstract:	Video synopsis facilitates the efficient analysis of surveillance footage by condensing extended video content into shorter segments and reorganizing moving objects temporally based on a predefined objective function. Various approaches have been developed to address challenges in video synopsis, such as object collisions, maintaining chronological order, reducing computational time, and managing scene complexity by grouping activities based on user-defined queries. However, many query-based detection techniques often overlook specific tasks like detecting intrusions, such as unauthorized entry, identifying individuals wearing unauthorized clothing, or detecting person substitution. This research extends one of the conventional query-based approaches by characterizing each object tube through its motion and visual features to enhance the detection of such intrusions. Moreover, it implements query-based video synopsis generation using a distributed architecture that utilizes multiple servers to boost performance. While the distributed model effectively mitigates and improves the query-based computational performance, it introduces potential security risks, including object tube substitution during cross-device transfers and the possibility of tube loss. The proposed system contributes by building a distributed system architecture that reduces computational complexity and provides a comprehensive solution that balances between efficiency, security, and reliability in handling surveillance video synopsis. It offers a two-layer security model; the first layer embeds a watermark within each object tube to safeguard against substitution during processing. The second layer serializes each object tube, ensuring no loss occurs during transmission, thus preserving the integrity of the entire process. This novel approach integrates distributed computing, watermarking, and serialization to enhance both the efficiency and security of the video synopsis process, providing a robust solution for large-scale surveillance video analysis. The simulation results indicate that the security module effectively extracts the watermark with a reasonable PSNR and identifies any tube loss. Additionally, the analysis reveals that utilizing a distributed system greatly enhances performance, which is critical for real-time applications. Specifically, the proposed distributed model reduces computation time by approximately 94% compared to sequential execution when implemented across twenty machines.
Keywords:	Video Synopsis; Video Surveillance; Distributed Computing; Object-Based Watermark; Tube Serialization
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	COMPARISON OF THE ENSEMBLE XGBOOST AND TRANSFORMER MODELS WITH MACHINE LEARNING FOR CLASSIFICATION OF INDONESIAN MUSIC MOODS OF THE 70'S AND 80'S ERA
Author:	SARTIKA LINA MULANI SITIO, THOYYIBAH.T, MARYANI, MAULANA ARDHIANSYAH, NENY ROSMAWARNI, YAN MITHA DJAKSANA, NUNIK DESTRIA ARIANTI
Abstract:	The aim of this research is to compare the Xgboost and transmormer machine learning models. In the field of psychology, research on musical mood aims to find out why humans have an emotional response to music. In the field of Music Information Retrieval (MIR), research on the emotion and mood of music aims to create music metadata to make it easier to manage and retrieve music as an entity. The method used in this research is the CRISP-DM method with the stages of business understanding, data understanding, data preparation, modeling, evaluation and deployment. This research has only reached the evaluation stage. This research used 1160 data using lyrics and audio data and used 31923 different words. This research uses a dataset of Indonesian songs from the 70's and 80's. The labels used for mood classification are happy, sad and neutral. The lyric representation used is word embedding and audio consisting of Chromagram, Mel-Spectrogram. This research uses the Xgboost model and transformer machine learning. With several analyzes used, the transformer model accuracy achieved is 98%.
Keywords:	Indonesian Music, Xgboost, Transformer, Crisp-DM
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	EMPIRICAL INVESTIGATIONS TO PREDICT STOCK PRICE USING REGRESSION TO IMPROVE THE TRENDS OF INORMATION TECHNOLOGY
Author:	G. RESHMA, SUDARSHAN TUMKUNTA, TGAYATHRI, T.V. HYMA LAKSHMI, N. NEELIMA, GARIGIPATI RAMA KRISHNA, BORRA BHAVITHA
Abstract:	Accumulating riches by astute investment—which doesn't! In actuality, numerous trading, financial, and even technological firms have been actively researching stock market movements and stock price prediction. Machine learning techniques have been used to generate a number of algorithms for stock price prediction. Here, we'll concentrate on mastering a number of well-known regression methods, such as support vector regression, regression tree, regression forest, and linear regression, and using them to solve this billion-dollar problem. The prediction's primary goal is to lessen the ambiguity surrounding investment decision-making. Following the random walk, the stock market suggests that the best forecast for tomorrow's value is today's value. Investors' beliefs are impacted by the stock market indices' extreme volatility. Because of the fundamental characteristics of the financial industry and, in part, the combination of known and unknown factors, stock prices are thought to be highly volatile and subject to abrupt fluctuations. There have been various attempts to predict stock price utilizing Machine Learning. Each research endeavour has a very different focus in three different ways. Target price changes might be short-term (tomorrow to a few days later), long-term (months later), or near-term (less than a minute). Less than ten specific companies, stocks in a specific industry, or almost all stocks can be included in the collection. A worldwide news and economic trend, specific firm attributes, or just time series data of the stock price can all be employed as predictors.
Keywords:	Stock, Regression, Trends, IT, Price.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	REVOLUTIONIZING INFORMATION RETRIEVAL: UNVEILING A NEXT-GENERATION AI-POWERED QUESTION-ANSWER SYSTEM FOR COMPREHENSIVE DOCUMENT ANALYSIS
Author:	ZAHER NAJWA, GHAZOUANI MOHAMED, CHAFIQ NADIA
Abstract:	Following the introduction of a new public procurement decree in Morocco, public institutions and businesses have faced challenges in comprehending the updated rules for awarding contracts. The complexity of the new decree has made it difficult for these entities to adapt to the changes, underscoring the necessity for detailed guidance and training. To address these issues, this article suggests creating an advanced system to analyze documents related to the new law, aiding employees by simplifying access to crucial information. Utilizing state-of-the-art text analysis and natural language processing, the proposed tool aims to enhance understanding and compliance with the decree by making legal information more accessible and easier to navigate. Different techniques were compared, namely, spaCy, Langchain, GloVe, BERT, and OpenAI Embeddings. In conclusion, our experiment has demonstrated that Langchain and OpenAI Embeddings surpassed the other techniques in terms of performance. Specifically, Langchain's specialized approach to text splitting proved to be exceptionally efficient in preprocessing documents for analysis, allowing for more nuanced segmentation that facilitated deeper understanding in subsequent processing stages. Similarly, OpenAI Embeddings offered superior capabilities in capturing the semantic richness of text, enabling our system to achieve higher accuracy and relevance in its responses.
Keywords:	Question-answer system, OpenAI, Langchain
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	MODEL OF UTAUT 2 FOR THE ADOPTION OF BLOCKCHAIN TECHNOLOGY IN IMPROVING QUALITY OF FINANCIAL REPORTING
Author:	DEVYLYA METRINESYA, FRISHELLA AMANDA YOKO, ANG SWAT LIN LINDAWATI
Abstract:	This study explores the adoption of blockchain technology to enhance the quality of financial reporting by utilizing the UTAUT2 model to assess the factors influencing its acceptance among internal accountants. Blockchain technology is essential to improve transparency, security, and accuracy in financial data. Despite the potential benefits of blockchain technology, the complexity of its application in financial records still requires further understanding. By applying a quantitative approach with data collected using questionnaires distributed to industry experts in the financial sector. The key findings reveal that most of the UTAUT2 factors have a positive significant effect on blockchain adoption, except for social influence that has a significant negative effect, along with hedonic motivation and price value do not have a significant influence on blockchain adoption. The conclusions of the study highlight the importance of technology-driven infrastructure development in financial reporting to present relevant, verifiable, comparable, understandable, and reliable financial statements. The rapid dynamics of the development of blockchain technology in various industries, including the financial sector, may limit the relevance of the findings in this study over time. Therefore, this study concludes by recommending exploring higher adoption by expanding the scope of future research.
Keywords:	Blockchain, Technology, Accounting, Financial Reporting, Distributed Ledger, UTAUT2
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	SECURE AND EFFICIENT MEDICAL IMAGE ENCRYPTION FOR MEDICAL CYBER PHYSICAL SYSTEM USING CHAOTIC WITH DNA SEQUENCE
Author:	KAILASAM SELVARAJ, KARTHEEBAN KAMATCHI
Abstract:	In contrast to the conventional healthcare industry, Medical Cyber-Physical Systems increasingly utilize smart sensor technology and health app to ensure enhanced healthcare. Medical imaging has become essential in recent years for disease diagnosis. These images include X-rays, ultrasound, and the brain which contains private and delicate information. There are many difficulties with the protected sharing and storing of medical images. The protection of medical information is increased through the use of encryption. The research recommends protected and lightweight medical image encryption using chaotic with DNA sequences for a medical cyber-physical system. First, the image is divided into blocks, rotated based on local pixel intensity variance, and shuffled using random permutation. The shuffled blocks are then reassembled, and a bitwise XOR operation completes the encryption, with a key generated from chaotic logistic maps and DNA symbols. Analytical evaluations show the proposed method outperforms earlier techniques, offering stronger security for medical data.
Keywords:	MCPS, Medical Image, Encryption, Chaotic, DNA, Xor operation, image blocks, image rotation, imaging, Measure and Integration
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	UTILITY LIST-BASED MINING AND RECURRENT NEURAL NETWORK FOR UTILITY ITEMSET MINING BASED TRANSACTIONAL DATA
Author:	G.N. SOWJANYA, M. BABU REDDY
Abstract:	The study of Frequent Itemset Mining (FIM) and High-Utility Itemset Mining (HUIM) is essential because it provides practical insights to improve company outcomes and explains customer behavior. To detect high-utility itemsets, including those with negative utilities, HUIM algorithms have been efficiently developed. Utility Itemset Mining (UIM) has changed into an important area of research aimed at identifying high-utility patterns in transactional databases, where item utility is measured using criteria such as profit, frequency, or relevance. This research introduces a new framework that associations Utility List-Based Mining (ULBM) with Recurrent Neural Networks (RNNs) for effective and scalable utility itemset mining in transactional data. The utility list organization stores significant utility data for each itemset, enabling direct and effective calculation of utility values without the need for candidate generation. This considerably reduces recollection usage and processing costs. To enhance prediction abilities and capture temporal relationships amongst transactions, we use an RNN, specifically a Long Short-Term Memory (LSTM) network, to detect sequential patterns in transactional data. This hybrid technique leverages the utility list's efficiency in performing utility computations while the RNN learns temporal correlations between itemsets and predicts high-utility itemsets across transaction sequences. Our experiments validate that the proposed method, RNN-UBLM, not only outperforms existing utility mining algorithms in terms of accuracy and effectiveness but also excels at capturing dynamic, time-varying utility patterns. This approach is particularly well-suited for requests in e-commerce, retail analytics, and other sectors where utility-based decision-making is critical.
Keywords:	Frequent Itemset Mining, High-Utility Itemset Mining, Utility List-Based Mining, Transactional data, Recurrent Neural Networks.
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	INSURECHAIN: A BLOCKCHAIN-BASED SYSTEM FOR SECURE, EFFICIENT AND INTEROPERABLE INSURTECH
Author:	EMAN YASER DARAGHMI, HAZIM HARB, YOUSEF AWWAD DARAGHMI
Abstract:	Insurance Technology (InsurTech) solutions aim to transform traditional insurance models to more personalized services, streamlined claims processing and faster delivery. However, records in InsurTech are fragmented and isolated, rather than interoperable and cohesive. The need for multiple access to insurance records had raised the interoperability challenges between clients and providers, which pose additional barriers to effective data sharing. Additionally, insurance data face increasing threats since several advanced techniques have been developed to violate digital privacy and security. Therefore, Blockchain, which is a distributed trusted immutable database solution and ledger technology, can solve these problems. Although several studies were proposed to employ the Blockchain for managing insurance records, there is still a need for more research to better understand, characterize and evaluate its utility in InsurTech systems. This paper proposes a Blockchain-based InsureTec system called InsureChain that provides interoperable, secure, and efficient access to the records by providers, clients and third parties while maintaining privacy. The InsureChain employs smart contracts for effectively managing the claims of clients, governing transactions, and monitoring computations through the enforcement of acceptable usage policies. For better efficiency, the InsureChain uses Proof of Authority (PoA) and an incentive mechanism that leverages the degree of provider’s nodes from the perspective of InsurTech systems by measuring their efforts regarding maintaining records and creating new blocks. The system was evaluated by reviewing privacy, integrity communication channels, and security and comparing consensus algorithms. The results show that the system achieves high privacy, integrity, and secure communication. The PoA outperformers the proof of work and proof of stake.
Keywords:	Insurance, InsurTech, Privacy, Confidentiality, Cybercrime, Blockchain, smart contracts
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text

Title:	PREDICTION OF POWER OUTPUT ON EXTERNAL COMBUSTION ENGINE USING REGRESSION MODELS
Author:	RIZQI FITRI NARYANTO, MERA KARTIKA DELIMAYANTI, RIZKY ADI, ABDURRAHMAN, PUTRI KHOIRIN NASHIROH, IMAM SUKOCO, FIQRI FADILLAH FAHMI, DAMAI YUDHA AKBAR EFFENDI, AFRILZA DAFFA NARYAPRAMONO
Abstract:	This study explores the application of machine learning regression models to predict power output in External Combustion Engine on Combined Cycle Power Plants (CCPPs) using a comprehensive dataset of 9,568 hourly observations from 2006 to 2011. Key ambient variables include temperature, pressure, humidity, and vacuum. To prevent overfitting, a 5x2 fold cross-validation strategy is employed, generating 10 unique training and testing sets. Several models are assessed, including Random Forest, XGB Regressor, Extra Trees, Hist Gradient Boosting, and LGBM Regressor. XGB Regressor demonstrates superior performance with a Mean Absolute Error (MAE) of 2.41 and Root Mean Squared Error (RMSE) of 3.37, making it the most accurate model. Additionally, the performance of ensemble models further highlights their reliability in predicting power output. The study emphasizes the importance of advanced machine learning techniques in optimizing power predictions, balancing computational efficiency, accuracy, and interpretability for large-scale industrial applications. Boosting Regressor provides a more equitable compromise between computational efficiency and performance, rendering it well-suited for implementations on a large scale. Furthermore, despite its marginally diminished accuracy, the Random Forest Regressor offers significant insights via the feature importance analysis, thereby augmenting interpretability. This study underscores the significance of sophisticated machine learning models in enhancing the precision and effectiveness of power output forecasts in CCPPs. It stresses balancing interpretability, computational cost, and accuracy in real-world applications.
Keywords:	Cross-Validation, External Combustion Engine, Machine Learning Models, Power Output, Performance Metrics
Source:	Journal of Theoretical and Applied Information Technology 31^th December 2024 -- Vol. 102. No. 24-- 2024
Full Text