		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

August 2023 | Vol. 101 No.16

Title:	DESIGN AND BUILD OF SEARCHING SYSTEM FOR THE NEAREST FISH SHOP ON AN ORNAMENTAL FISH MARKET WEBSITE USING THE HAVERSINE ALGORITHM
Author:	WAWAN GUNAWAN, EMIL R. KABURUAN, BIKTRA RUDIANTO, ANGGI PUSPITASARI4, BUDI SUDRAJAT, DIRGAHAYU ERRI, ARIEF NUR HIDAYAH
Abstract:	To address the problem of high shipping costs and long delivery times due to the distance between stores and rare stocks of fish, we have developed a system on Aqua Store ID that allows buyers to find the nearest fish shop based on their address. The system utilizes data from the Aqua Store ID internal database, which includes store data and fish product data. Additionally, we use the haversine algorithm to calculate the distance between the user's coordinates and the store coordinates. The data used in this study is obtained from the Aqua Store ID internal database. It includes information about the stores and their respective fish products. By utilizing the haversine algorithm, we can calculate the distance between the user's location, obtained from the address they input, and the coordinates of the stores in the Aqua Store ID database. To ensure the reliability and functionality of the system, thorough testing was conducted using the black box method. The black box method involves testing the system's functionality without examining its internal code. Predefined steps were followed during the testing process, and the results indicate that the system runs well and successfully provides search results for the nearest fish shop based on the user's address. With the implementation of this system, we expect that buyers on Aqua Store ID can easily locate the nearest fish shop based on their address. This will help solve problems related to shipping costs and delivery times, especially if the buyer is able to find a fish shop that is closer in proximity. By reducing the distance between the store and the buyer, we provide a more efficient and cost-effective buying and selling experience for ornamental fish on Aqua Store ID.
Keywords:	Haversine Algorithm, Shortest Distance Measurement, Buying And Selling Ornamental Fish, Waterfall Method, Black Box Method)
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	IMPLEMENTATION OF AUTOMATIC NUMBER PLATE RECOGNITION TO DETECT BAD DEBT VEHICLES
Author:	LUSIUS EVODIUS TANZANIA, ANDRY CHOWANDA
Abstract:	Automatic Number Plate Recognition is very useful in various institutions. Researchers continue to develop and improve vehicle number plate detection in various countries. Each vehicle number plate has a different pattern in each country. There are various stages in the introduction process so that the vehicle number plate can be recognized. In recognizing each character, various algorithms can be used. Convolutional Neural Network (CNN) is a model that can find objects. Many researchers previously performed image classification using CNN. Many previous studies conducted research on handwriting in various types of languages. The research to develop a suitable CNN model to recognize the characters on vehicle license plates in Indonesia. The existing dataset is divided into 36 classes with 0 - 9, A - Z, and each class is approximately 1016 images in the grayscale form. In training the CNN model, the Python programming language and Keras library were used to make it easier to create layers on the CNN model. The evaluation and validation process will use a confusion matrix to find out the results of the predictions for each class. In this training, we will use LeNet and VGG16 architectures. Based on these two architectures, the best results will be selected to be used in the character recognition process on vehicle number plates. The LeNet architecture yielded 98% with 50 epochs during the training process, while VGG16 was 84% with 20 epochs.
Keywords:	CNN, ANPR, Indonesian Plate Number, Indonesian Plate Number Recognition, Image Pre-Processing
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	IMPLEMENTATION OF TRANSFER LEARNING MOBILENETV2 ARCHITECTURE FOR IDENTIFICATION OF POTATO LEAF DISEASE
Author:	TIKA ADILAH M, DINAR AJENG KRISTIYANTI
Abstract:	Potatoes are one of the third most important food crops in the world. Potato farming has problems in the form of diseases that attack the leaves. These diseases can affect the quality of potato plants, resulting in crop failure. Digital image processing is a method that can be used to assist farmers in identifying potato leaf diseases. The development of digital image processing has been carried out, one of which is by using the Convolutional Neural Network (CNN) algorithm. CNN requires big data. CNN architecture will experience overfitting if it uses little data, where the classification model has high accuracy on training data but poor accuracy on test data. This research utilizes Transfer Learning and Augmentation methods to avoid overfitting on too little data. Transfer Learning method used in this research is MobileNetV2. The results of the trials in this study indicate that the MobileNetV2 Transfer Learning method has good classification performance results and produces a high accuracy value of 99.6%.
Keywords:	Convolutional Neural Network, Leaf Disease Classification, MobileNetV2, Potato Leaf Disease, Transfer Learning
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	CLOUD DATA SECURITY USING CRYPTOGA AND BLOCKCHAIN RECOVERY
Author:	S.SUDHA, DR.S.S.MANIKANDASARAN
Abstract:	Most companies face the risk of a data breach revealing customers and employees stored personal information. Over time, the occurrence of such events has increased and can result in significant costs for the organization concerned. The key goal of this paper is to identify these problems and investigate possible solutions to the process of sensitive data handling. Proposed cloud data storage mainly considers data security for privacy and data recovery for unexpected data losses. To provide these features, our proposed framework includes CryptoGA for data encryption and decryption, Blockchain with erasure coding technology for data storage and backup.
Keywords:	Cloud, CryptoGA Data Encryption and Decryption, Data Storage and Backup.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	PRAGMATIC INVESTIGATIONS TO SMART DUSTS LOCATION APPRAISAL PRECISELY USING MACHINE LEARNING
Author:	MR. TATA BALAJI1, KURRA UPENDRA CHOWDARY, DR. P VENU MADHAV, DR.A GEETHA DEVI, MRS.T. MAHALAKSHMI, DR.SURYA PRASADA, RAO BORRA6 N.JAYA
Abstract:	Rough terrain that is difficult or impossible to access does not lend itself well to traditional Wireless Sensor Networks (WSNs). Smart dust is a technology that gathers remote sensing data from harsh terrain by utilising a network of numerous microscopic sensors. The small sensors are dispersed in large numbers across difficult terrains using airborne distribution from drones or aeroplanes, eliminating the need for manual placement. Although it is clear that this technology can be applied to a wide range of remote sensing applications, the small size of smart dusts essentially precludes the integration of complex circuitry on tiny sensors. This poses a number of challenges, one of which is locating the smart dusts. In order to pinpoint the precise location of events detected by the smart dusts, this study suggests a localization algorithm. General regression neural network is used in the method to forecast the locations. Because real smart dusts aren't readily available, we created a simulator to assess the proposed method's accuracy when used to monitor forest fires. The simulation trials indicate that the method is reasonably accurate.
Keywords:	Smart Dust, Machine Learning Algorithms, IOT, Sensor Applications
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	THREE-STEP BLOCK METHODS WITH DIFFERENT OFF-STEP POINTS INTERVAL FOR SOLVING SECOND ORDER INITIAL VALUE PROBLEMS
Author:	KAMARUN HIZAM MANSOR, OLUWASEUN ADEYEYE, ZURNI OMAR
Abstract:	Block methods are numerical methods adopted for the direct solution of higher order ordinary differential equations (ODEs) with no need for reduction to a system of first order ODEs. Hybrid block methods, which combines the use of both on-step and off-step points in the derivation of the block method, are known for performing better in terms of absolute error in comparison with block methods developed using only on-step points. However, when considering multistep methods, the off-step step points can be selected at different intervals, and it is important to know which interval choice for the off-step points gives the best results. This article considers a three-step block method with selection of two off-step points at all three intervals, which leads to developing three different methods. The first method is the three-step block method with off-step points selected within the first interval, while the second and third methods are three-step block methods with off-step points selected in the second and third intervals respectively. These resultant methods are used to solve second order initial value problems and comparison is made among the three block methods, and with existing studies. The three hybrid block methods (HBMs) showed comparable performance among themselves but displayed better accuracy when compared to existing studies in terms of absolute error. The basic properties of the HBMs were also tested and they were found to be zero-stable, consistent, and hence convergent.
Keywords:	Hybrid Block Method, Numerical Solution, Second Order, Initial Value Problems
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	ASSESSMENT OF THE ENTERPRISE ARCHITECTURE BASED ON THE PATTERN
Author:	JIHANE LAKHROUIT, KARIM BAÏNA
Abstract:	The enterprise architecture (EA) is the organizing logic for business processes and IT infrastructure, reflecting the integration and standardization requirements of the company’s operating model. Enterprise architecture (EA) is an approach to managing the complexity of an organization’s structures, information technology (IT), and business environment. This paper presents a complete pattern-based methodology for analyzing the complexity of enterprise architecture. The objective is to propose an evaluation methodology for guiding designers and architects in evaluating and improving the EA models. The methodology measures the mico-view and the macro-view metrics. Furthermore, our enterprise architecture patterns system will be used for automated support to manage the evaluation of enterprise architecture complexity.
Keywords:	Enterprise Architecture; EA patterns; Analysis of Enterprise Architecture; Complexity; Heterogeneity.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	REVOLUTIONIZING FOETAL CARDIAC ANOMALY DIAGNOSIS: UNLEASHING THE POWER OF DEEP LEARNING ON FOETALECHO IMAGES
Author:	DIVYA M O, M S VIJAYA
Abstract:	The use of Artificial Intelligence (AI) has amplified in various fields, with remarkable results in medicine in recent times. Despite the potential of AI in the medical field, there are still many unexplored areas due to data unavailability. One such area is cardiac foetal anomaly diagnosis, which is poorly diagnosed globally with a rate of only 50%. The complexity of the task requires a high level of expertise to understand minute hints and conduct thorough exams for accurate image captures. In this research, the FoetalEcho_V01 dataset was used for foetal cardiac anomaly diagnosis, consisting of pre-classified ultrasound images representing 15 different anomalies and a class representing normal heart images. The deep learning models which are efficient in producing potential classifiers for ultra sound scan images are identified. The models are CNN, AlexNet, VGG16 and ResNet50. The best performing deep learning models were used to produce classifiers, and their performance was evaluated. The results showed that the deep learning models performed well on the FoetalEcho_V01 dataset images for diagnosing structural cardiac anomalies in the foetus, with consistent performance as demonstrated by the calculated standard deviation. The results obtained from the research for the FetalEcho_V05 dataset are as follows. The CNN model achieved a precision of 0.94, recall of 0.89, accuracy of 0.90, and F1 score of 0.91. Comparatively, the AlexNet model demonstrated a precision of 0.92, recall of 0.87, accuracy of 0.89, and F1 score of 0.89. The VGG16 model exhibited precision of 0.91, recall of 0.85, accuracy of 0.87, and F1 score of 0.88. Lastly, the ResNet50 model displayed a precision of 0.93, recall of 0.90, accuracy of 0.93, and F1 score of 0.93. Among these models, the CNN model emerged as the best classifier for the FetalEcho_V05 dataset, with its superior performance in terms of precision, recall, accuracy, and F1 score.
Keywords:	Cardiac Heart Defect, Classification, Deep Learning, Diagnosing Foetal Cardiac Anomalies, Prenatal Diagnosis
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	ANALYSIS OF ECC AND ZKP BASED SECURITY ALGORITHMS IN CLOUD DATA
Author:	E.JANSIRANI, DR.N.KOWSALYA
Abstract:	Cloud computing is now used to store massive amounts of data in various areas such as business, military universities, and so on. We can obtain data from the cloud based on the user's request. Clients use cloud storage services to upload their files as well as authentication information to a cloud storage server. Cloud server (CS) must demonstrate to a verifier that he is actually storing all of the client's data unchanged in order to ensure the availability and integrity of the data that is being saved for clients. To save data on the cloud, many obstacles must be surmounted. There are several methods that can be taken to address these issues. For data protection, cryptography methods are gaining popularity. In cloud computing, a singular method is ineffective for providing high-level data security. In this paper, we establish a new security method based on Hybrid Elliptic Curve Cryptography and Zero Knowledge Proof (HECCZKP) Algorithm, which is used in this suggested system to provide data protection when compared to Elliptic Curve Cryptography (ECC) and Zero Knowledge Proof (HECCZKP).
Keywords:	Cloud Computing, Data Security, Elliptic Curve Cryptography, Zero Knowledge Proof, Discrete Logarithm Problem
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	MULTI-USER REAL-TIME SIGN LANGUAGE RECOGNITION OF WORDS USING TRANSFER LEARNING OF DEEP LEARNING NEURAL NETWORKS
Author:	RASHMI GAIKWAD, LALITA ADMUTHE
Abstract:	The hearing and speech impaired people use sign language for communication. Nevertheless, other people cannot understand sign language. A real-time American Sign Language recognition system will help to reduce this gap of communication. This paper presents a solution for real-time sign language recognition of words in American Sign Language. In this research transfer learning of two deep learning pre-trained modules available in the Tensorflow object detection repository namely SSD MobileNet V2 FPNLite 320x320 and SSD ResNet50 V1 FPN 640x640 is implemented. The dataset consisting of signs of eight words is generated exclusively for this research. Signs obtained from a single user are used for training and testing of the networks but real-time detection can be done on signs performed by multiple users. A comparison of performance of the two networks is also done for the same dataset. Accuracy in terms of Confidence level is 100% for same signer detection and for different signers it comes in between 60% to 80%.The precision and recall of SSD MobileNet V2 came to be 91% and 71% respectively while that of SSD ResNet50 V1 came to be 87% and 74% respectively.
Keywords:	Deep Learning Neural Networks, Sign language recognition (SLR), SSD MobileNet V2 FPNLite 320x320, SSD ResNet50 V1 FPN 640x640, Convolutional Neural Network (CNN), Dataset.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	A WAVELET BASED STATISTICAL TECHNIQUE FOR DENTAL CARIES SEVERITY CLASSIFICATION USING FUZZY FEATURE SELECTION
Author:	PRITI SEHGAL, ROLI BANSAL, PRERANA SINGH
Abstract:	Dental caries is a disease resulting in tooth decay due to bacterial infection in it. The caries can be classified into four severity types i.e., incipient, moderate, advanced and severe, depending on the depth or extent to which they affect the tooth. In this paper, we have classified the caries into different severity types using wavelet domain-based feature extraction methods followed by fuzzy feature selection. The input dental X-ray images of caries infected tooth are subjected to feature extraction using three statistical methods i.e., Wavelet Completed Local Binary Pattern (WCLBP), Wavelet Completed Local Binary Count (WCLBC) and Wavelet Completed Local Ternary Pattern (WCLTP). Since severity/extent of caries is subjective in nature, we further propose a fuzzy rule-based system for selecting important features from the extracted ones, suitable for severity classification. This also addresses the dimensionality problem faced by the above methods. The improved results of the proposed model, using fuzzy feature selection, prove its potency. This study justifies that WCLTP followed by fuzzy rule-based feature selection and Adaboost classification proves to be one of the most effective ways to identify and classify the dental caries based on the severity.
Keywords:	Dental Caries, Medical Imaging, Fuzzy Feature Selection, Severity Based Classification
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	DEEP LEARNING DRIVEN (DLD) PROSTHETIC HAND GESTURE RECOGNITION AND OBJECT TRACTION FOR DISABLED PERSON THROUGH SURFACE EMG(sEMG)
Author:	SURYA.S, RAMAMOORTHY.S
Abstract:	People affected with Neuro diseases and lost hands in accidents unable to perform their activities by their own, the technology aided supportive devices which accelerate the activities in day today activities may be the primary requirement for this type of semi-paralyzed people. This research paper presents a novel approach to predicting prosthetic hand gestures using machine learning and deep learning techniques. Surface electromyography (sEMG) signals are collected from the user's forearm muscles, which are then processed to identify the intended hand gesture. The proposed model contributes the people affected semi- paralyzed stage to achieve their intended activities through Deep learning based object detection model. The dataset consists of seven hand gestures commonly used in daily activities. To establish a baseline performance, the K-Nearest Neighbor (KNN) algorithm is employed and achieves an accuracy of 96%. To improve the prediction accuracy further, a Convolutional Neural Network (CNN) model is developed and trained on the same dataset. The CNN model achieves an accuracy of 86%, which is lower than the KNN model but still demonstrates promising results. In addition to the hand gesture prediction model, an object detection model is also developed. The dataset for this model is created from scratch and consists of images of everyday objects. The model uses a combination of deep learning techniques to identify the object in the image and assigns a corresponding gesture that can be performed with the object using the prosthetic hand. The proposed models have several potential applications in the field of prosthetics. They can be used to develop prosthetic devices that are more intuitive and responsive to the user's intended gestures, improving their overall functionality and user experience. Moreover, the object detection model can be extended to identify more complex objects and gestures, expanding the range of activities that can be performed using the prosthetic hand. This study shows that it is possible to correctly predict prosthetic hand gestures using machine learning and deep learning techniques. The proposed models are a significant contribution to the field's research because they exhibit encouraging findings and have a number of possible applications in the prosthetics industry. The findings have implications for the creation of prosthetic hand control systems that are more dependable and precise and that can be used in everyday life. Overall, this study shows how machine learning and deep learning techniques could advance in the field of prosthetics and, eventually, enhance the quality of life for people who have lost limbs.
Keywords:	Prosthetic Hand, Gesture recognition, EMG Signals, Deep learning Model, Convolutional Neural Network.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	REAL-TIME ANOMALY DETECTION IN INTERNET OF THINGS DEVICES USING TEMPORAL CONVOLUTIONAL NETWORK
Author:	GHAZOUANI MOHAMED, ABDERAHMANE DAIF, ARDCHIR SOUFIANE, MOHAMED AZZOUAZI
Abstract:	Anomaly detection is the examination of specific data points and the detection of rare occurrences that appear suspicious because they are different from the established pattern of behaviors. Anomaly detection is nothing new, but the increase in data volume makes manual tracking impossible. When such an anomaly occurs, it is sometimes difficult to realize it, and the delay between the beginning of the anomaly and its observation can be a day or more, depending on the case. This article proposes a neural network-based model for real-time anomaly detection in Internet of Things sensors. The aim of this study is to detect defective Internet of Things Devices at the Laboratory Information Technology and Modeling, Faculty of Sciences Ben M'sik, Hassan II University Casablanca, Morocco. Different neural network models were compared, namely, Long Short-Term Memory (LSTM), gate recurrent unit (GRU) and a Temporal Convolutional Network (TCN). In conclusion, our experiment has demonstrated that the TCN model surpasses other models in terms of performance. The impressive performance of these models reaffirms the significance of this approach and its potential for enhancing preventive maintenance of Internet of Things devices.
Keywords:	TCN, LSM, GRU, Anomaly Detection, Internet of Things
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	DEEP NEURAL SYSTEM FOR IDENTIFYING CYBERCRIME ACTIVITIES IN NETWORKS
Author:	YERININTI VENKATA NARAYANA, Dr. MOORAMREDDY SREEDEVI
Abstract:	Nowadays, the increase in communication networks increases the risk of cybercrime. These crimes affect all individuals, from children to adults. Cybercrime causes serious impacts that many nations were researching various detection frameworks and safeguard approaches such as regulation of internet usage, establishing an organization to deal with cybercrime issues and adaptive forensic techniques. The various limitations regarding the cybercrime detection framework must be eliminated. Hence this current article reviewed different cybercrime detection frameworks based on deep neural models. Here several literature works were discussed with their advantages and their limitations. Furthermore, in the performance analysis section, the results of the few works were compared. Subsequently, common defeats and their reason were explained in the discussion part. Finally, future works have directed the following studies to improve the efficiency of the detection frameworks.
Keywords:	Cybercrime, Intelligence Frameworks, Classification, Features.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	USABILITY ANALYSIS OF HUMAN COMPUTER INTERACTION IN GOOGLE CLASSROOM AND MICROSOFT TEAMS
Author:	HENOCH JULI CHRISTANTO, STEPHEN APRIUS SUTRESNO, ANDREW DENNY, CHRISTINE DEWI
Abstract:	As online learning continues to gain prominence, the usability and user satisfaction of digital platforms play a crucial role in facilitating effective and engaging learning experiences. This study focused on evaluating the usability and user satisfaction of two popular online learning platforms, Microsoft Teams and Google Classroom. By assessing various variables such as learnability, efficiency, memorability, errors, and satisfaction, the study aimed to provide valuable insights into the user experience of these platforms. The study employed a quantitative approach, gathering data through surveys and measurements from a sample of users who had experience with both Microsoft Teams and Google Classroom. The participants were asked to rate their satisfaction levels, provide feedback on usability aspects, and share their overall perceptions of the platforms. The results of the study indicated that both Microsoft Teams and Google Classroom received positive feedback, the consecutive values for Learnability, Efficiency, Memorability, Error, and Satisfaction in Microsoft Teams are 47.25%, 29.67%, 34.27%, 21.39%, and 25.20%, respectively. On the other hand, in Google Classroom, the values are 49.48%, 30.69%, 35.71%, 18.94%, and 26.45%, respectively. However, Google Classroom demonstrated higher satisfaction ratings compared to Microsoft Teams, suggesting a stronger preference among users. It excelled in user-friendliness, intuitive design, and navigation, while Microsoft Teams faced occasional usability challenges despite its collaborative features. This study highlights the importance of usability in online learning platforms and its impact on user satisfaction. Platforms like Google Classroom, with superior usability, tend to generate higher user satisfaction due to their intuitive interfaces and seamless interactions. Conversely, platforms like Microsoft Teams can enhance their usability to address user concerns and improve the overall user experience. The study provides valuable guidance for educators, developers, and designers, emphasizing the need for continuous user research, usability testing, and iterative design to enhance platform usability and create engaging learning environments.
Keywords:	Usability, HCI, online learning platforms, Google Classroom, Microsoft Teams
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	A NOVEL OPTIMIZATION-BASED APPROACH FOR HARMONIC ASSESSMENT IN HYBRID MICROGRIDS
Author:	DEVIKA RANI MOTUKURI, P.S.PRAKASH, M.VENU GOPALA RAO
Abstract:	Hybrid microgrids have shown great promise in recent years as a way to deal with the difficulties of combining renewable energy sources with conventional power plants in contemporary power grids. However, the performance and stability of such microgrids might be negatively impacted by the existence of harmonic disturbances. Effective control solutions are required to reduce harmonic distortions and guarantee reliable power transmission. This research suggests a Proportional-Integral-Derivative (PID) controller based on Gray Wolf Optimization (GWO) be used to evaluate harmonics in hybrid microgrids. Inspired by the pack dynamics of grey wolves, the GWO algorithm is a robust and simple optimization method with quick convergence rates. This research aim is to optimize the PID controller settings in real-time to reduce harmonic distortions in the microgrid system by combining the GWO method with the PID controller. A thorough simulation model of a hybrid microgrid, including multiple distributed energy supplies, energy storage devices, and loads, is used to assess the effectiveness of the suggested control technique. We compare the GWO-based PID controller to traditional PID controllers and various optimization approaches like particle swarm optimization to demonstrate its harmonic reduction ability. Due to enhanced harmonic evaluation, the GWO-based PID controller increases the power quality and decreases the harmonic distortion, according to simulations. This controller is ideal for hybrid microgrids since it adapts to changing system conditions and load needs. This work shows that the GWO-based PID controller can efficiently mitigate harmonics and improve renewable-integrated power system performance, contributing to hybrid microgrid control strategies.
Keywords:	Renewable Energy Sources, Power Quality, Gray Wolf Optimization, Particle Swarm Optimization, PID Controller, Total Harmonic Distortion
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	GRASSHOPPER OPTIMIZATION-BASED NEUTROSOPHICAL FUZZY CONVOLUTIONAL NEURAL NETWORK FOR ENHANCED MOVING OBJECT DETECTION
Author:	S SARAVANAKUMAR, M. LINGARAJ
Abstract:	Detecting moving objects is a cornerstone of computer vision research and has many practical uses in security, robotics, video analysis, and virtual reality. This paper presents a novel approach, the Grasshopper Optimization based Neutrosophical Fuzzy Convolutional Neural Network (NFCNN), for enhanced moving object detection. The proposed approach integrates the Grasshopper Optimization Algorithm (GOA), neutrosophic principles, and fuzzy logic into a Convolutional Neural Network (CNN) architecture to improve moving object detection accuracy, robustness, and efficiency. The GOA is employed to optimize the parameters of the NFCNN, enabling adaptive learning and feature extraction from input data. Neutrosophic principles are integrated into the NFCNN to handle uncertain and imprecise information, capturing the nuances and contradictions in moving object detection. Fuzzy logic is incorporated to manage the imprecision and uncertainties inherent in object detection tasks. The proposed GOA-NFCNN is evaluated on benchmark datasets, and existing practices are compared to the outcomes. The experimental results demonstrate the superiority of the Grasshopper Optimization-based Neutrosophical Fuzzy Convolutional Neural Network’s accuracy, robustness, and computational efficiency. Integrating GOA, neutrosophic principles, and fuzzy logic in the NFCNN yields significant improvements in moving object detection. The proposed approach enhances the ability to handle complex motion patterns, occlusions, and variations in lighting conditions, resulting in more accurate and reliable object detection in dynamic environments.
Keywords:	Moving Object Detection, Grasshopper Optimization Algorithm, Neutrosophical Fuzzy Logic, Convolutional Neural Network (CNN), Enhanced Object Detection, Computer Vision
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	THE ROLE OF IT GOVERNANCE AS MULTIGROUP MODERATION IN THE RELATIONSHIP OF ORGANIZATIONAL DEMOGRAPHICS, EXTERNAL ENVIRONMENT CHARACTERISTICS, AND ORGANIZATIONAL CULTURE TO FRAUD
Author:	AGUNG BUDIWIBOWO, ENDANG SITI ASTUTI, MUHAMMAD SAIFI, MOHAMMAD IQBAL
Abstract:	This study focuses on companies related to fraud and the purpose of this study is to reveal the causal relationship between the influence of exogenous variables, namely Organizational Demography, External Environmental Characteristics, Organizational Culture, and IT Governance as moderating variables to the endogenous variable, namely fraud, so this study uses explanatory research. The population of this study are companies that go public (offering business ownership to the general public) which are listed on the Indonesia Stock Exchange until 2020 and commit fraud in the company's financial reporting or within a certain period of time do not report financial statements, namely 216 companies. The sampling technique used in this study was purposive sampling with the determination of the sample size using the 5% slovin formula. This research uses Structural Equation Modeling (SEM) analysis with Partial Least Square (PLS) method. The result fo this study show that Organizational Demographics does not have a significant positive effect on Fraud; In IT Governance with a low category, External Environmental Characteristics does not have a significant effect in a positive direction on Fraud, but in high category, External Environmental Characteristics have a significant effect in a positive direction on Fraud; and Organizational Culture has a significant negative effect on Fraud. The originality of this research is the disclosure of the impact of the moderating variable of IT Governance on the influence of Organizational Demographics, External Environmental Characteristics, and Organizational Culture on Fraud. In this study, the moderating variable of IT Governance is a categorical variable which is divided into two groups, namely the low group and the high group.
Keywords:	Information Technology Governance; Organizational Demographics; Organizational Culture; External Envirnoment Characteristics; Fraud
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	AUGMENTED REALITY IN HIGHER EDUCATION, A 10-YEAR SYSTEMATIC LITERATURE REVIEW
Author:	ABDULLAH AHMED ELNAQLAH, NURULLIZAM JAMIAT, TAMER NAZIR MADI
Abstract:	Higher education is increasingly using augmented reality (AR) as a teaching and learning tool. The purpose of this systematic literature review (SLR) is to investigate the current situation of the art of research on augmented reality (AR) in higher education. The study synthesizes 20 peer-reviewed articles that were written between 2011 and 2022 with a focus on examining certain criteria, such as the total number of studies performed over time, the countries that have used augmented reality in higher education, the duration of the study, the academic fields that have used augmented reality, the variables that have been measured, and the data collection techniques used. The study comes to the conclusion that augmented reality technology has the potential to improve student outcomes in higher education, including engagement and learning processes. The review identifies the knowledge gaps that might be investigated in future studies and offers insightful information about the use of augmented reality in higher education overall.
Keywords:	Augmented Reality, Higher education, AR, Mobile learning, Education.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	IMPROVED EMAIL MARKETING DECISION MAKING IN A CHURN PREDICTION CONTEXT USING MACHINE LEARNING ALGORITHMS
Author:	REDOUAN ABAKOUY, EL MOKHTAR EN-NAIMI, ANASS EL HADDADI, LOTFI ELAACHAK
Abstract:	Several studies have been carried out to study the problem of churn in different fields such as telecommunications and marketing. In fact, the literature studied is based on the classic approach of machine learning as well as deep learning. Thus, we have noticed that research works related to the study of the phenomenon of churn in the field of digital marketing are limited. Customer churn in email marketing suggests that the opportunity for both leads and potential sales are wasted, but the nightmare of any digital company is when the customers unsubscribe from the campaign list without any form of advice. Digital companies may find it hard to respond and take corrective actions to increase their profitability and revenue which is the highest priority in any business, including digital ones. To overcome these kinds of issues digital companies should take a proactive approach and identify potential customers before they leave. Thanks to the data available in several data sources, customers’ transactions including purchasing habits can be extracted and then analyzed later. In this perspective of research, we present in this article a study on some of the most applied machine learning algorithms in this context, challenging customer churning difficulty by predicting their behaviors during several email marketing campaigns. In this work, we applied the model by utilizing Machine Learning algorithms. The greatest model in our study reached a predictive accuracy of 82%, measured by the F1-Score.
Keywords:	Digital Marketing; Machine Learning; Churn; Email Marketing; Prediction Models
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	ENHANCING CREDIT RISK MANAGEMENT IN THE BANKING SECTOR THROUGH MACHINE LEARNING-BASED PREDICTIVE MODELS
Author:	ALI K. ABDUL RAHEEM, MAYS ZUHAIR
Abstract:	Credit risk plays a vital role in the functioning of the banking sector, as banks extensively engage in providing loans, credit cards, mortgages, and other financial products. However, the increasing number of credit card users has led to a rise in credit card default rates, posing challenges for banks. To address this issue and effectively control credit risk, leveraging data analytics becomes crucial. This study aims to predict loan defaults, enabling banks to proactively mitigate potential losses by offering alternative options to borrowers. To achieve this objective, we propose a system that utilizes various machine learning classification algorithms. Specifically, we explore and compare the performance of Logistic Regression, XGBoost, k-Nearest Neighbors, Neural Network, and Random Forest models in predicting credit risk. Our findings reveal that the Random Forest model demonstrates exceptional accuracy, with a forecast accuracy of 98% in assessing the credit risk of credit card users.
Keywords:	Credit Risk; Machine Learning; Prediction; Risk Management; Deep Learning.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	OPTIMIZED PRACTICAL BYZANTINE FAULT TOLERANCE (O-PBFT) ALGORITHM USING GROUPING METHOD FOR CONSORTIUM BLOCKCHAIN
Author:	JANNAH YUSOFF, ZARINA MOHAMAD, HASNI HASSAN, AZNIDA HAYATI ZAKARIA@MOHAMED, SHADI A ALJAWARNEH
Abstract:	Blockchain is a distributed ledger that records every transaction that has ever occurred in a system. A consensus algorithm is required in blockchain to ensure that all transactions function properly. At present, the consensus algorithm that is commonly used in consortium blockchain is Practical Byzantine Fault Tolerance (PBFT) algorithm. PBFT requires all nodes in the network to participate in the consensus process. However, PBFT still has some problems such as communication overhead, low throughput, and high latency due to increasing the number of nodes in the network. To overcome these disadvantages, an optimized PBFT (O-PBFT) algorithm is proposed. The O-PBFT algorithm used grouping method in a mid-stage (prepare stage) to reduce communication complexity and assign random Byzantine nodes to improve consensus efficiency. The consistency protocol in O-PBFT were modify from the original PBFT so that O-PBFT can reach consensus with less communication in a stable network. Experimental results show that O-PBFT algorithm reduces number of communication times between nodes, increases transaction throughput, and improves consensus efficiency compared to the original PBFT algorithm. Experimental results prove that O-PBFT algorithm can be used when many nodes are involved.
Keywords:	Consensus Algorithm, Practical Byzantine Fault Tolerance (PBFT), Blockchain, Consortium Blockchain, Grouping Method.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	DIPLOMA AUTHENTICATION MADE EASY WITH BESU BLOCKCHAIN: A NOVEL PLATFORM FOR SECURE VERIFICATION
Author:	MOHAMED GHAZOUANI, ABDERAHMANE DAIF, TARIK BOURAHI, MOHAMED AZZOUAZI
Abstract:	Educational institutions, employers, and students have long struggled with the challenge of authenticating diplomas. We have identified that the absence of a dependable and secure system to verify the legitimacy of diplomas has resulted in the proliferation of counterfeit credentials and fraud. This problem is further exacerbated by the worldwide nature of education and the lack of uniformity across different regions and countries. To tackle this issue, a novel blockchain-based platform has been created for diploma authentication. The platform, built on the Besu blockchain, offers a tamper-proof and secure system for verifying the authenticity of diplomas. Advanced cryptographic techniques are employed to ensure that the data is unchangeable and can be authenticated by authorized parties in real-time. The creation of this platform comes at a critical juncture, as the need for trustworthy and secure diploma authentication has become increasingly pressing. With the rise of online education and remote work, it is crucial to have a system in place that can guarantee the authenticity of credentials. This new platform is not only secure but also free, making it accessible to everyone. It has the potential to revolutionize the way that diplomas are authenticated and could have far-reaching implications for the future of education and employment.
Keywords:	Blockchain, Authentication of Diplomas, Hyperledger Besu, Smart Contract, Educational Certificate
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	HYBRID APPROACH BASED ON MACHINE LEARNING AND DYNAMIC CASE BASED REASONING FOR COVID-19 CLASSIFICATION
Author:	MOHAMED KOUISSI, ISMAIL BENSASSI, EL MOKHTAR EN-NAIMI, ABDELHAMID ZOUHAIR
Abstract:	COVID-19 has reached almost every country and affected millions of people because of its incredible speed of spread. Covid-19 had a bad territorial impact in its different dimensions: health, economic and social. Predicting infected cases in advance can help reduce the rate of its transmission and reduce the number of infected people. In this article, we propose a hybrid approach of covid-19 cases classification based on Multi-Agent System (MAS), Dynamic Case Based Reasoning approach, and the support vector machine supervised machine learning algorithm. The aim of this approach is to classify the cases’ diagnostic as infected by COVID-19 or not to estimate suspected cases. It consists of an audio classifier for cough recognition by the SMV algorithm in the retrieve step of the Dynamic Case Based Reasoning to fetch a set of similar cases to our target case. To increase and enhance our working dataset, the data augmentation technique has been used. Then, our data has been preprocessed and cleaned to ensure the consistency and converted into a Mel-Spectrogram to extract useful features and transform them into significant feature images, so they will be used as input to our model. The test set of our model has reached an accuracy of 80% and a precision up to 95%.
Keywords:	Dynamic Case Based Reasoning (DCBR), Covid-19, Support Vector Machine (SVM), Mel-Spectrogram, Multi Agents System (MAS)
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	ASPECT-BASED SENTIMENT ANALYSIS ON CHATGPT IN TWITTER USING BIDIRECTIONAL ENCODER REPRESENTATIONS FROM TRANSFORMERS (BERT)
Author:	HANDRIZAL, ANANDHINI MEDIANTY NABABAN, RICKY ALAN
Abstract:	ChatGPT, a chatbot developed by OpenAI, is currently being widely discussed, especially on social media platforms like Twitter. Many users have provided positive feedback regarding this chatbot. However, some have provided negative feedback. To understand the sentiment of users regarding ChatGPT, this research conducted aspect-based sentiment analysis using the Bidirectional Encoder Representations from Transformers (BERT) model. The aspects analyzed include general aspects, functions, performance, and potential of ChatGPT. The data used in this research was obtained from the social media platform Twitter. The BERT model used is IndoLEM/IndoBERT-base-uncased, which has been pre-trained on 220 million Indonesian language words from various sources. From the conducted tests, the model was able to achieve an f1-score of 84% for the general aspect, 89% for the functional aspect, 98% for the performance aspect, and 98% for the potential aspect.
Keywords:	Sentiment Analysis, Aspect Based Sentiment Analysis, ChatGPT, BERT, Twitter
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	COMPARATIVE STUDY AND PROPOSAL TO USE ENHANCED BELLMAN-FORD ALGORITHM FOR FASTER PATH COMPUTATION ON SDWAN CONTROLLER
Author:	MOHIT CHANDRA SAXENA, MUNISH SABHARWAL, PREETI BAJAJ
Abstract:	Software-Defined Wide Area Network (SDWAN) technology has revolutionized network management by providing efficient and reliable communication across geographically dispersed locations. One critical aspect of SDWAN is path computation, which determines the optimal routes for data transmission between network nodes. Traditional shortest path algorithms like Bellman-Ford, Dijkstra, and SPFA are commonly used for path computation. However, the increasing scale and complexity of SDWAN networks demand more efficient algorithms. In this research paper, we propose an enhanced shortest path algorithm called SCBF (Shortest Path Computation Based on Bellman-Ford) specifically designed for SDWAN controllers. SCBF incorporates optimization techniques to reduce computational overhead and improve runtime efficiency. By building upon the principles of the Bellman-Ford algorithm, SCBF introduces novel optimizations that expedite path computation. To evaluate SCBF's performance, we conduct a comparative study against traditional shortest path algorithms. Through extensive simulations using various network topologies and traffic scenarios, we demonstrate that SCBF outperforms the traditional algorithms in terms of runtime efficiency. SCBF achieves faster path computation on the SDWAN controller, reducing computational complexity and improving scalability. The comparative study showcases the advantages of SCBF in terms of reduced network latency, improved throughput, and enhanced scalability. The findings contribute to the development of more efficient path computation algorithms for SDWAN controllers, enabling faster decision-making in network routing and resource allocation. These improvements lead to enhanced performance and reliability in SDWAN deployments. Future research and real-world implementations can explore the practical implications of SCBF in SDWAN environments, further validating its effectiveness. SCBF has the potential to provide significant benefits in terms of reduced network overhead and improved network management, ultimately enhancing the overall SDWAN experience.
Keywords:	SDWAN, Shortest path, Path Computation, Algorithm, SDN, link performance, Bellman Ford, Dijkstra, smart WAN
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	EXPLORING THE IMPACT OF LIVE STREAMING FOR E-COMMERCE BUSINESS: A SYSTEMATIC LITERATURE REVIEW
Author:	RINTIS EKO WIDODO, TOGAR ALAM NAPITUPULU
Abstract:	Live streaming has become increasingly prevalent in the realm of e-commerce, offering businesses a unique platform to interacting with customers in real-time. The utilization of live streaming, initially popularized on social media platforms, has now gained significant traction in the realm of e-commerce. Research in this field has been conducted by researchers from various countries. In this study, the impact of live streaming on e-commerce business is explored using a systematic literature review. By examining a range of scholarly articles and studies, this review identifies key findings regarding the impact of live streaming on various aspects of e-commerce, such as consumer behavior, consumer engagement, purchase intention, and social presence. The analysis reveals that live streaming, through its interactive and immersive nature, enhances consumer engagement, trust, and purchase intention. Moreover, it demonstrates that effective utilization of live streaming can bolster brand awareness, foster customer loyalty, and drive sales. These findings highlight the significant role that live streaming plays in shaping consumer decision-making processes and its potential as a powerful marketing tool for e-commerce businesses.
Keywords:	E-Commerce, Consumer Engagement, Live Streaming E-Commerce, Purchase Intention, SLR
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	QURAN RECITER IDENTIFICATION: TECHNIQUES AND CHALLENGES
Author:	MOHAMMED ALATIYYAH
Abstract:	Speech-based intelligent systems are gaining an increasing popularity and importance due to their wide range of applications in our daily life. Most of the research efforts within speaker identification target the English language. However, efforts that target the Arabic language and the holy Quran are still limited. For Muslims, the Holy Quran is the main religious book of Islam. The Holy Quran verses must be read according to very restricted rules known as “Tajweed’ to guarantee the correct pronunciation of verses. The task of identifying the holy Quran reciter or reader based on many features in the corresponding acoustic wave is known as Quranic reciter identification process. It is considered a more challenging task than other speaker identification tasks as it depends on “Tajweed”. As a result, this paper provides a survey of the Holy Quran reciter identification problem, describing the proposed techniques, models and challenges in this area, focusing on the advances made during the last decade to help future researchers who aim to enhance previous results.
Keywords:	Reciter Identification, Quran, Arabic Language, Speaker Identification, Machine Learning.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	PREDICTION MODEL BEHAVIOR BURNOUT AMONG MANAGERS AND EMPLOYEES OF START-UPS WITH ARTIFICIAL INTELLIGENCE
Author:	NOS SUTRISNO, ANDRE HASUDUNGAN LUBIS, MARISCHA ELVENY1, LORENA NAINGGOLAN, MAYANG SEPTANIA IRANITA, RAHMAD SYAH
Abstract:	Employee burnout is a common problem in start-up businesses, and it has a negative impact on the productivity and performance of employees and managers. This study proposes a prediction model based on the Support Vector Machine (SVM) and Gradient Boosting algorithms to identify potential cases of burnout among managers and employees of start-up companies. The model predicts the likelihood of burnout based on data on work-related factors such as workload, job demands, and job control. The performance of the SVM and Gradient Boosting models was evaluated using real-world datasets from startups in predicting burnout cases, according to the results, The proposed model can help managers identify and treat burnout cases early on, resulting in improved employee well-being and overall performance. with the results where age, work experience, job type greatly affects fatigue in the world of work with a sensitivity of 0.91 and a specificity of 0.92.
Keywords:	Prediction; Burnout; Support Vector Machine (SVM); Gradient Boosting; Artificial Intelligence
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	CONSTRUCTION OF THE OPENING MECHANISM OF THE MIMOSA PUDICA BASED ON THE ANALYSIS OF VIDEO FRAMES FOR ARTIFICIAL LIFE IN CONFINEMENTS
Author:	JULIO CESAR BAUTISTA ROSAS, MARIA FERNANDA GUILLEN PEREZ, RODOLFO ROMERO-HERRERA
Abstract:	The behavior of plants and trees, even when conducted unconsciously, we cannot ignore the fact that very useful algorithms can be obtained. In this project, mechanisms based on the movement and behavior of plants (mimosa pudica, Venus flytrap, toad, fruit trees, etc.) that can interact with humans were developed; for example, relieving stress situations when human beings are in states of confinement, such as the one experienced by the COVID-19 pandemic. This is because in many hospitals and closed places it is not feasible to introduce vegetation due to allergies or that the living system may be affected. In the project, digital image and video processing were applied to observe the movement of the plant. Key point detection techniques will be used for the digital processing of frames. With the data obtained, the behavior of the plant or tree is simulated for which the basic mechanics of the plant are built, and sensors are adapted. An additional advantage is that it requires a minimum of maintenance and has an impact on digital health. In the project, digital image and video processing were applied to observe the movement of the plant. For the digital processing of frames, key point detection techniques were used, a practice that, unlike other methods, simplifies the analysis and implements segmentation such as geodesy that are not used for this type of analysis, but that simplify the process and improve movement detection. With the data obtained, the behavior of the plant is simulated for which the basic mechanics of the plant are built, and sensors are incorporated to detect the environment surrounding the plant. An additional advantage is that it requires a minimum of maintenance and has an impact on digital health, with a mathematical model obtained through frames.
Keywords:	Mimosa Pudica, Behavior, Plants, Trees, Intelligence.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	PROCESS MINING APPROACH FOR DISCOVERING AND ANALYZING THE HEALTHCARE PROCESSES IN PYTHON
Author:	ABDEL-HAMED MOHAMED RASHED, NOHA E. EL-ATTAR, DIAA SALAMA ABDELMINAAM, MOHAMED ABDELFATAH
Abstract:	As healthcare expands, there is an increasing need for innovative approaches to improve healthcare quality. One such approach is process mining which involves data analytics to extract insights from careflows and patient data in real-time. The process mining techniques can support business processes based on event log from information systems to model those processes and uncover challenges and bottlenecks; it supports the professionals of the field with a view of the problems that are currently occurring in the field. This paper applied process mining methods and techniques in the healthcare field to analyze the careflows and gains meaningful insights. The proposed approach has three main steps; preprocessing the dataset, applying miner algorithms to discover the process model, and applying the performance analysis to discover the existence deviation in the discovered model. This work can contribute to bring a Python (PM4Py framework) as a process mining tool rather than traditional tools that typically neglect to help process mining methods in large-scale analysis. This proposed method can be applied across several hospitals; it analyzes the patient careflows from the control-flow and the performance perspectives. It provides accurate analysis and insights to hospital administrators. It furthermore, it provides accurate analysis and insights to hospital administrators. The paper used a dataset of cardiology patients in an Egyptian hospital. The results of the applied approach based on PM4Py framework are efficient and satisfactory; it gives objective analysis and insights to hospital managers to improve performance care processes.
Keywords:	Process mining, Healthcare process, Cardiology, PM4Py, Event log.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	FACE RECOGNITION MODEL FOR ATTENDANCE SYSTEMS: A SYSTEMATIC LITERATURE REVIEW
Author:	BRIAN WIJAYA AQRALDO, MEYLIANA
Abstract:	Technological developments are currently very rapid in all industries. Technology has also entered our daily lives, for example when taking attendance, both at school and at work, we already use a sophisticated attendance system. Most of the existing attendance systems use fingerprints and RFID cards, which of course will cause many errors and system failures, as well as fraud in attendance. To overcome this, an attendance system was developed using facial recognition technology. FR technology itself is a science that comes from artificial intelligence which of course plays with the algorithm model in the system. This makes attendance systems developers confused to choose which algorithm model is good to implement on attendance systems. Therefore, this paper aims to find trends in the development of attendance systems using face recognition, the best algorithm for face detection and face recognition, and the best algorithm for face liveness detection. This can be answered by using a Systematic Literature Review with the PRISMA model approach, where data on papers were published in 2007-2023. From a review of the literature, it was found that there were the most trends in the development of attendance systems for face recognition in the education industry for student attendance, the best algorithm for face detection was MTCNN , the best algorithm for face recognition in attendance systems was CNN with an accuracy rate of 98.87 % Based on a literature review, and the best algorithm for face liveness detection is the ECT algorithm, because it is the most stable for liveness detection.
Keywords:	Face Recognition Algorithm, Face Detection Algorithm, Face Liveness Detection Algorithm, Attendance Systems, Systematic Literature Review
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	A NOVEL WOSRCNN-BASED TRUST MODEL WITH SECURE ROUTING AND DATA TRANSMISSION IN WSN USING CLF_AVOA AND ASCII-DSAES
Author:	MAHAMMAD MASTAN, G. JAI ARUL JOSE, LOUAY A. HUSSEIN AL-NUAIMY
Abstract:	Wireless Sensor Networks (WSNs) are networks in which to detect and gather data from the environment, it utilizes various types of Sensor Nodes (SNs) positioned in a specific location. Therefore, for transferring the data accurately as of source to destination, it requires secure Data Transmission (DT). It is prone to error since the data moves in the in-secured channel from one SN to another. Energy-efficient data gathering is also challenging owing to the limited energy resources of each SN. Thus, in this work, to accomplish an optimal trade-off betwixt security and resource utilization, by utilizing Chebyshev Levy Flight African Vulture Optimization Algorithm (CLF_AVOA) and American Standard Code for Information Interchange (ASCII) to Decimal Sorting adapted Advanced Encryption Standard (ASCII-DSAES) algorithms, a novel Weight Optimized Softplus Relu Convolutional Neural Network (WOSRCNN)-centric trust model with Secure Routing (SR) and DT is proposed. To determine the node details, a Node Discovery Message (NDM) is primarily transmitted after the SNs are positioned in the required environment. Then, it extracts the node features. After that, for the detection of trusted nodes, trust scores are calculated utilizing WOSRCNN. Then, by utilizing Euclidean Distance (ED) along with CLF_AVOA, distance and optimal paths are selected for the trusted nodes. With the ASCII-DSAES mechanism, the sensed information is partitioned and encrypted; then, it is transferred to the Base Station (BS) where the partitioned data is combined and stored in the server for further usage. The experimental results displayed that in contrast to the conventional frameworks, the presented model provides high security and throughput with minimum delay.
Keywords:	Wireless Sensor Network (WSN), Optimal path selection, Convolutional Neural Network (CNN), Secure routing, Advanced Encryption Standard (AES), Data partitioning, Trust identification.
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	USING CLUSTER ANALYSIS FOR REVEALING GENDER EQUALITY PATTERNS IN EU ICT EDUCATION AND EMPLOYMENT
Author:	VOLODYMYR TOKAR, DMYTRO TYSHCHENKO, TAMARA FRANCHUK, VALENTYNA MAKOIEDOVA, ANDRII LOTARIEV
Abstract:	Women are underrepresented in all aspects of ICT, including education and employment. This issue can be attributed to a lack of encouragement and opportunities for women to pursue ICT careers, as well as implicit biases and stereotypes leading to women being overlooked for roles in the industry. Cluster analysis is a useful tool for analyzing and grouping data to identify patterns and trends. This article aims to investigate the state of gender equality in information and communication technology education and employment across the European Union member-states using cluster analysis. The study uses data from the Eurostat to determine the female to male ratios of new bachelor entrants studying information and communication technologies, as well as employees working in the ICT in EU member-states. The article distinguishes gender equity leaders, adopters and laggards among EU member-states. The data was organized into six clusters of EU member-states with a focus on gender equality among those who enter into bachelor-level ICT programs and are employed in the field. Moreover, the article examines the ratio of female to male enrolled in information technology in Ukraine analyzing the case of the State University of Trade and Economics. It was found that the ratio of female to male does not have a stable upward trend. The article proposes some initiatives aimed at increasing the number of girls and women in ICT. This study provides valuable insights into the current state of gender equality in the ICT sector across EU member-states. The findings of this study can inform policymakers and stakeholders in developing targeted interventions to address gender inequality in the ICT sector. This will promote a more inclusive and diverse ICT workforce, which is essential for economic growth and social development.
Keywords:	Cluster Analysis, EU member-states, Female Empowerment, Gender Equality, Information Communication Technology, Gender Diversity Management
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	HYBRID LOSS SENSITIVITY FACTOR AND MUTATED ANT LION OPTIMIZER FOR OPTIMAL DISTRIBUTED GENERATION PLACEMENT WITH MULTIPLE LOADINGS
Author:	NUR ATIQAH BINTI ABDUL RAHMAN, ZULKIFFLI BIN ABDUL HAMID, ISMAIL BIN MUSIRIN, NUR ASHIDA BINTI SALIM
Abstract:	In power system planning and operation, the main objective is to deliver consistent electrical power at low power losses and stable voltage magnitude. One of the ways in which this objective can be accomplished is by incorporating Distributed Generation (DG) into distribution network. DG uses small-scale technologies to generate electricity close to consumers. Improper planning of DG invites several unwanted consequences which can lead to unstable system and inefficient use of electrical energy. Thus, placement of DG units at the right location with adequate size is of necessary. This paper proposes a new method for DG placement using hybrid Loss Sensitivity Factor (LSF) and Mutated Ant Lion Optimizer (MALO). The incorporation between LSF technique and optimization algorithm for optimal DG placement, in addition to the development of a new hybrid MALO algorithm are the contributions of this study. The objectives of this study are; (1) to improve voltage stability in distribution network using optimal DG placement and; (2) to develop a new hybrid optimization algorithm for optimal solution and fast computation. The proposed technique was tested under various loading conditions to see its robustness. Several experiments and comparative studies have demonstrated that the proposed hybrid LSF-MALO technique has successfully minimized real power losses while ensuring a stable voltage stability condition in the system. In terms of the optimization performance, the proposed LSF-MALO managed to offer a reasonable computation with optimal solution concurrently.
Keywords:	Distributed Generation, Loss Sensitivity Factor, Mutated Ant Lion Optimizer, MALO, Loss Minimization, Voltage Stability
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	A NOVEL METHOD TO ESTIMATE HIDDEN NEURONS IN ENSEMBLE FLOOD FORECASTING MODEL
Author:	NAZLI MOHD KHAIRUDIN, NORWATI MUSTAPHA, TEH NORANIS MOHD ARIS, MASLINA ZOLKEPLI
Abstract:	Machine learning model such as neural networks have been widely adopted to provide flood forecast. Self-adaptability in neural networks enable them to learn pattern by their own and adjusted the connection between the neurons, but still producing error. In many neural network adaptations for flood forecasting, the number of hidden neurons is normally randomly selected which can caused overfitting problem in the network. In this study, a novel method to estimates the hidden neuron is proposed to overcome this problem. This method integrates the evaluation of various convergence theorem criteria with grid search to estimates the hidden neuron. By having this integration, optimal number of fix hidden neurons can be determined. This method is used in the ensemble model that based on neural networks to forecast the water level based on rainfall data. Based on the performance measurement using Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Nash Sutcliffe Efficiency (NSE), it is found that the integration of convergence theorem and grid search can be used to fix the number of hidden neuron and has reduce the error that led to overfitting in the forecast.
Keywords:	Neural network, Hidden neuron estimation, Flood forecasting, Grid search, Convergence Theorem
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text

Title:	SHIP TRAJECTORY PREDICTION FOR ANOMALY DETECTION USING AIS DATA AND ARTIFICIAL INTELLIGENCE : A SYSTEMATIC LITERATURE REVIEW
Author:	ARIF BADRUDIN, SISWO HADI SUMANTRI, RUDY AGUS GEMILANG GULTOM, I NENGAH PUTRA APRIYANTO, HENDY RISDIANTO WIJAYA5, INDRAJANI SUTEDJA
Abstract:	According to reports from the Maritime Security Agency (Bakamla), Marine and Fisheries Ministry (KKP), and Ministry of Transportation, many illegal ship activities have occurred in 2022, such as dropping anchors at the wrong time and place, which indicates illegal fishing, not turning on AIS while in Indonesian territory, violating ALKI limits (Indonesian Archipelagic Ocean Currents); and ship activities that are of the greatest concern are the activities of transferring oil cargo (ship-to-ship) in Indonesian territory without going through ports. This is not only detrimental but can also endanger the security and safety of shipping in the Indonesian Sea area. Relevant parties need to find out about these violations, provide evidence of violations, and warn. This study aims to find the right model based on artificial intelligence so that it can detect ship violations earlier and produce evidence of violations in Indonesian waters. This study uses a systematic literature review (SLR) of written sources from Scopus, IEEE Explorer, and Google Scholar with the keywords AIS data, trajectory prediction, and anomaly detection to find solutions that have been done before. This systematic review explores the current state of research on ship trajectory prediction for anomaly detection using Automatic Identification System (AIS) data and artificial intelligence. AIS is a widely used system for vessel tracking that collects information such as ship location, speed, and direction. This method produces 559 related papers, of which only 28 are appropriate and form the basis of this research. The results show that AIS data and artificial intelligence can help find violations through ship trajectory prediction using DBSCAN. This research can be further developed by exploring AIS data and completing it with comparisons with speed, heading, and travel time which will produce activity predictions where they stop.
Keywords:	Ship Trajectory Prediction, Automatic Identification System (AIS) Data, Systematic Literature Review (SLR), Anomaly Detection, Machine Learning, Artificial Intelligence (AI).
Source:	Journal of Theoretical and Applied Information Technology 31^st August 2023 -- Vol. 101. No. 16-- 2023
Full Text