		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

March 2023 | Vol. 101 No.5

Title:	THE ROLE OF EXPERIENCE AND SELF-EFFICACY IN THE TECHNOLOGY ACCEPTANCE MODEL: A STUDY CASE FOR PRACTITIONERS AND TRAINEES
Author:	QUYNH NH, HOAI NT, LOI NV
Abstract:	The purpose of the current study is to estimate the factors stimulating the acceptance and use of information technology (IT) systems by practitioners and trainees. To solve this problem, the research has applied the technology acceptance model (TAM) - this model has been becoming one of the widely accepted research models in the field of IT applications. However, when conducting a comprehensive review of existing literature, it was found that the relationship results in TAM were inconsistent between the studies given the diversity of predictive factors in the TAM along with cultural differences. Therefore, this study attempts to suffice this gap by assessing the interaction factors in TAM with the e-learning context of working adults. The results show that experience, enjoyment, and self-efficacy affected learners' perceived usefulness and ease of use, which in turn affected the satisfaction and intent of the behavior. Therefore, this study has provided insight into how learners shape their attitudes and intentions for e-learning.
Keywords:	E-Learning, Practitioners And Trainees, Experience, Self-Efficacy, TAM
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	A STRATEGY FOR SELECTIVE DYNAMIC FIREWALL AND PROACTIVE SECURITY METHODS ON CLOUD
Author:	S REKHA GARIKAMUKKALA, DR. V.RAVI SHANKAR
Abstract:	The most crucial concern of modern cloud computing industry is the security. The security for traditional systems can be provided with tightly coupled components such as firewalls or tremendously strong authentication methods. But in the case of cloud computing, the extra delay caused by security mechanisms that are more complicated is not quite adaptable. The recent research outcomes have demonstrated significant improvements in terms of detection of the attack, however those methods are either formulated as reactive methods to reduce the time complexity or highly time complex if the methods are proactive. However, the security concerns must be addressed. In the recent past, a good number of researchers have aimed to solve this long persistent challenge. These outcomes from the parallel research, are highly time complex and cannot cater to cloud computing service level agreement demands. Thus, the proposed work analyzes the possibilities of reduction of model or framework complexities by introducing highly time efficient reduction method using the proposed correlation and further reduces the complexity of the firewalls using proposed regression method. The proposed framework establishes the proactive security framework with 93% accuracy in characteristics reduction, 25% reduction in analysis time and 99% accuracy in attack preventions.
Keywords:	Collective Correlation, Dimensionality Reduction Regression, Dynamic Firewall, Rule Generator, Firewall Deployment
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	TOWARDS A FULLY CLOUD-BASED PLATFORM FOR ARABIC NLP
Author:	MOHAMMED NASRI , SARA CHAOUKI , MOSTAFA SAADI, NOREDINE GHERABI
Abstract:	Cloud Computing is getting more and more debated in the IT industry today. Its evolution is leading the next generation of internet services. Natural Language Processing (NLP) is a field of Artificial Intelligence that focuses on understanding, manipulating and generating human language by machines. Thus, the NLP is really at the interface between computer science and linguistics. It, therefore, concerns the ability of the machine to interact directly with humans. Arabic NLP is very poor compared to other languages such as English or German due to the complexity of this language and the lack of resources. In this work, we propose a new system for Arabic NLP based entirely on the Cloud. This system is based on two steps. It firstly uses a bridge between Arabic and other developed languages (English for occurrence) and then uses of the already developed features for that language. Hence those features apply not on the Arabic text but instead on the translated (the English) one. In some cases, the result needs to be in Arabic, in which case, we use the bridge another time to translate English result into Arabic. This can either be used in real NLP systems, such as Translation, IR, QA, Sentiment Analysis, or for validation or comparison purposes, especially for those who work in NLP and use other approaches. Experiments have been performed on a prototype we developed and the results obtained are satisfactory for this first version.
Keywords:	Arabic NLP, Cloud-based platform, SaaS NLP
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	TOWARDS AN EFFICIENT IOT SYSTEM BY INTEGRATING BLOCKCHAIN IN IOT
Author:	RAJAT VERMA, NAMRATA DHANDA, VISHAL NAGAR, MUDRIKA DHANDA
Abstract:	The extension of the Internet of Things (IoT) started with the invention of the Electromagnetic Telegraph in 1832. Since then, IoT has never looked back and continues to grow rapidly. Today, IoT is capable enough of interacting with billions of users together, generating a voluminous amount of data by performing different operations on IoT devices by the users. Data and Operations are precious assets of an organization that must be kept secure. It should have Confidentiality, Integrity and Availability, which denotes the normal CIA triad in cyber-security. With this, the 3 A’s (i.e., Authentication, Authorization and Auditing) should also be followed for keeping the facts and figures secured. Alternatively, the negative side of the development in the technological field is also evolving i.e., cyber-attacks. Today with the development, cyberattacks are much stronger, so conventional security must be upgraded with some SMART solutions. The reason is that the old solutions are somehow responsible for the bugs and concerns it generates in the IoT systems. Here, Blockchain comes into the picture, which is a network technology, which follows immutability, transparency and decentralization. Blockchain emerged as the next big thing when it was used with Bitcoin as a security measure. This Paper illustrates the integration of Blockchain in IoT for enhancing the security aspects of IoT systems. Moreover, this paper also analyses the efficiency of Blockchain transactions in IoT, making a normal IoT device into a Blockchain-secured IoT device (IoT-B). A variety of use cases are available, but the authors have considered a General-Purpose Input Output (GPIO) emulator as an IoT device for examining the efficiency of blockchain transactions.
Keywords:	Blockchain, IoT, IoT-B, Privacy, Security.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	AN INTEGRATED QUANTUM AND BIOMETRIC KEY GENERATION BASED CLOUD DATA SECURITY FRAMEWORK FOR STRUCTURED AND UNSTRUCTURED ELECTRONIC HEALTH RECORDS
Author:	NAGABABU GARIGIPATI, DR. V. KRISHNA REDDY
Abstract:	Most of the conventional cloud data security and authentication protocols are independent of dynamic key generation with variable size user integrity on large homogeneous electronic health records (EHRs) due to memory and time computations. Also, biometric based key generation and management plays an essential role for cloud data security and user authentication process in real-time applications. In the unstructured electronic health records, biometric based data security provides a strong integrity verification and multi-level security on large data size in cloud computing environment. In this work, a hybrid biometric key generation based encryption and decryption framework is implemented using the integrated quantum transformation on large heterogeneous databases. In this work, homogeneous 2D and heterogeneous unstructured 3D EHRs along with clinical data are used to compare the proposed model with the state of art algorithms. Experimental results show that the proposed quantum biometric based multi-level cloud data security framework has better data security than the conventional models in terms of avalanche affect and average runtime (ms) computations.
Keywords:	Biometric Key Generation, Multi-Authority Security, Cloud Computing, Heterogeneous Medical Records
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	INTEGRATION CLUSTER AND PATH ANALYSIS BASED ON SCIENCE DATA IN REVEALING STUNTING INCIDENTS
Author:	MAMLUATUL MARCHAMAH, ADJI ACHMAD RINALDO FERNANDES, SOLIMUN, NI WAYAN SURYA WARDHANI, HENIDA RATNA AYU PUTRI
Abstract:	The purpose of this study is to applied big data to explore the factors that influence the prevalence of stunting in Wajak, to model these factors using cluster analysis models and integrated path analysis, and to develop an information system that can provide comprehensive information about stunting in the Wajak. This study uses a descriptive and explanative approach, namely using Discourse Network Analysis, cluster analysis, path analysis, and integration of cluster and path analysis. The sample of this research is children under five in Wajak District who were selected using stratified random sampling. The distance measure that has the highest model goodness vsquaredue R^2in modeling using the integration of cluster analysis with path analysis is the Mahalanobis distance measure. The cluster analysis with Mahalanobis distance produces 3 clusters where cluster one is a toddler who has a low stunting category, cluster two is a group of toddlers who has a moderate stunting category, and cluster three is a group of toddlers who has a high stunting category. The originality of this study is the application of Discourse Network Analysis to obtain new variables followed by a comparison of three distances namely euclidean, manhattan, and mahalanobis in modeling using cluster integration and parametric paths.
Keywords:	Big Data, Stunting, Integration Cluster, Path Analysis
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	METHODS, IMPROVEMENT AND CHALLENGES ON CLOTH AND HAIR ANIMATION
Author:	NG HUI XIN , RAJERMANI THINAKARAN
Abstract:	Animation may create the illusion that something is moving by taking a succession of drawings and displaying them one after another. This phenomenon is known as the illusion of movement. The generation and control of movement is a significant obstacle in both hair and clothing animation. The purpose of this study is to investigate the techniques utilized in hair and clothing animation, as well as how these techniques have changed over time and what obstacles they currently face. For this study, a systematic literature review (SLR) method was used, and 47 manuscripts from various databases between the years of 1987 and 2022 were found. According to the study, continuum model and clump mode were used for hair animation while geometric model and physical model were used for clothing animation. As technology has advanced, improvements in seams, wrinkles, blonde hair, and shading have made it easier for animators to animate clothing and hair. The animated item or character also looks better and more realistic as a result of this improvement. The precision, estimation of human shapes and sizes utilizing various devices and various fabric materials are the final obstacles for clothing animation, although it is still challenging to produce realistic hair animation in interactive applications.
Keywords:	Cloth Animation, Hair Animation, Methods, Improvement, Challenges
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	SECURITY THREATS, COUNTERMEASURES AND DATA ENCRYPTION TECNHINQUIES ON THE CLOUD COMPUTING ENVIROMENT
Author:	ABRAR ALISMAIL, ESRA ALTULIHAN, RAWAN BUKHOWAH, MOUNIR FRIKHA
Abstract:	In the world of computing, cloud computing has emerged as one of the fastest-emerging technologies. As more and more businesses store data in the cloud, it's possible to hack, alter, or delete that data since the servers are in remote locations. The main problem with cloud computing is security breaches, which must be solved flawlessly. Since the use of cloud computing is increasing, more security issues are emerging. Therefore, it is necessary to study this topic, analyze the new emerging security issues targeting cloud computing data, and search for solutions to mitigate these issues. It is important to know what types of attacks target the cloud computing data, what techniques are used for such attacks, and how to protect cloud computing users from these security issues. Cloud computing data security can be implemented in a variety of ways, one of which is through data encryption. This paper reviews the major data security issues present in cloud computing. A comparison study was made of several encryption techniques used in cloud computing. Moreover, this paper discusses the encryption platforms available for safeguarding cloud data and compares them to others based on factors such as security, complexity, usability, supported OS, advantages, and disadvantages
Keywords:	Cloud, Threats, Mmitigation, Techniques, Data Security, Encryption.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	MEASURE THE LEVEL CAPABILITY IT GOVERNANCE IN EFFECTIVENESS INTERNAL CONTROL FOR CYBERSECURITY USING THE COBIT 2019 IN ORGANIZATION: BANKING COMPANY
Author:	FRANCISKUS XAVERIUS ADRIAN, GUNAWAN WANG
Abstract:	Nowadays, Companies must be able to use information technology effectively and efficiently in order to support company goals in the face of rapid technological advancements. Banking is one of the industries that must use IT to compete with banks and other financial technology firms. This research conducted to assess the level of capability IT Governance which is the method that the Author used for this research is COBIT 2019 Framework and NIST Cybersecurity for point of view about cybersecurity. The purpose of this research is to make readers know about Domain COBIT 2019 and Cybersecurity that can be used to optimize IT process and cyber security. In this study, qualitative data was gathered through interviews and observations, while quantitative data was gathered through questionnaires based on COBIT 2019 Framework controls. Furthermore, an analysis of the current and expected IT capability levels was performed, and recommendations were made to help the company achieve the expected capability level. Furthermore, the NIST cybersecurity framework is being used as an additional recommendation for companies in terms of cybersecurity and preventing cybersecurity attacks. The Selected COBIT Process in This Research are APO07 (Managed Human Resources), BAI07 (Managed IT Change Acceptance and Transitioning), BAI09 (Managed Assets), DSS01 (Managed Operations), DSS02 (Managed Service Requests and Incidents), DSS03 (Managed Problems), DSS05 (Managed Security Services.). The result of the capability levels of IT governance processes at company are at level 2 and level 3 and Based on the calculations that the Author had done, the score of capability rate in this IT function is 2.28 and the company has a target for future is to reach max level 5, that about having a Gap Score 2.72. As a result, a recommendation is made to improve these processes by referring to the best practices suggested by COBIT and in accordance with company needs. There are also some recommendations from the NIST Cybersecurity framework so the company is aware of data protection in an era where all systems are interconnected and integrated with one another. Making applications that provide LMS and Incident/IT Service Recording, as well as analysing trends that occur in the company for incidents/services, monitoring and reporting on a regular basis, and creating a policy, procedure, and team that handles cybersecurity within the company, are the main recommendations.
Keywords:	COBIT, Framework, Internal Control, Banking, Cybersecurity
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	ASSESSING THE INFLUENCE OF READINESS TOWARDS THE CONTINUOUS USE INTENTION OF FUND TRANSFER FINTECH IN INDONESIA
Author:	SOVIA DUMIYANTI, SFENRIANTO
Abstract:	Financial Technology has become increasingly popular because it provides various solutions to its users. One of them is reducing transaction costs of basic banking and money transfer services. However, despite the advantage offered and the prominent market opportunity, many products in this market have been unsuccessful and ceased operation. This high rate of failure in the market prompts Fintech to know its customers better. As a fund transfer fintech, Flip must understand its customer's continuous use intention to ensure its success. Therefore, this research employs the TRAM model to analyze the factors that influence the continuous use intention of fund transfer applications. This study utilizes the Technological Readiness and Acceptance Model, which constructs includes Optimism, Innovativeness, Discomfort, Insecurity, Perceived Ease of Use, and Perceived Usefulness. Research results on 402 samples showed that Continuous Use Intention is positively and significantly influenced by Perceived Ease of Use and Perceived Usefulness. Optimism and Innovativeness are shown to positively and significantly affect Perceived Ease of Use. Perceived Ease of Use is positively and significantly influenced by Optimism but negatively influenced by Insecurity.
Keywords:	Continuous Use Intention, Interbank Transfer, Flip, Financial Technology, TRAM Model
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	AN EFFECTIVE IDENTIFICATION AND ANALYSIS FOR BRAIN TUMOR DIAGNOSIS USING AN EFFICIENT MACHINE LEARNING TECHNIQUE
Author:	REDROUTHU VAMSI KRISHNA, K V V SATYANARAYANA
Abstract:	The surgeon would be assisted in clinical diagnosis by machine learning diagnostic image detection, but the effectiveness of the detection sample depends on a huge amount of labelled data. The most effective and accurate strategy to prevent the epidemic of brain tumor disease was to diagnose it, given the disease's rapid rise. In earlier studies, reducing the input data's dimension resulted in a reduction in data storage space. Therefore, a novel Tiger-based Support Vector Machine (TbSVM) technique was proposed for diagnosing brain tumor disease. Initially, this system performed preprocessing, feature extraction, prediction, and segmentation for diagnosing brain tumors. Here, utilizes the wiener filter to eliminate the noise during the preprocessing stage, and the GLCM feature to extract the features. Finally, the prediction and segmentation process were preceded with the help of a designed model to provide an accurate result for diagnosing the brain tumor. The suggested model was compared against other methods in the results section using several metrics.
Keywords:	Tiger Algorithm, Optimization, Feature Extraction, Wiener Filter, Machine Learning.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	ENHANCING ARABIC FAKE NEWS DETECTION FOR TWITTERS SOCIAL MEDIA PLATFORM USING SHALLOW LEARNING TECHNIQUES
Author:	ALBARA AWAJAN
Abstract:	One of the most significant and severe issues facing social media nowadays is fake news, especially in social media applications - the intentional deception of users through the spread of propaganda, rumors, or misleading information about a variety of people or issues. A very common social media platform with the highest rapidly increasing number of users in the Middle East is Twitter. However, along with an increase in the number of Twitter users has come an increase in the amount fake news being spread across the platform. This, in turn, has caught the attention of researchers who seek nothing more than an online experience free of fake news. As are result, through utilizing both the transformer-based language and recurrent neural network models, this paper introduces an intelligent classification model that identifies fake news presented in Arabic tweets. Afterwards, this research presents a comparative study between deep learning and shallow learning. In Addition, to enhance the effectiveness of the proposed model, researchers also built an Arabic Twitter dataset that included 206,080 tweets. Results have shown that pre-trained deep learning models performed much better than shallow models when it came to identifying fake news in Arabic. That is, a 95.92% accuracy rate was reached when researchers applied shallow learning to the pre-processed data set; however, after applying the LSTM model to the same data, the accuracy rate increased to 96.71% while after apply the Bert model to that same data, the accuracy rate increased to a near perfect 99%.
Keywords:	Fake news, Shallow Learning, Deep learning, Twitter, Social Media.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	IMPLEMENTATION OF DIGITAL TECHNOLOGIES IN PERSONNEL MANAGEMENT SYSTEM OF ENTERPRISES OF ROCKET AND SPACE INDUSTRY
Author:	ALEXEY AKIMOV, ALEXEY TIKHONOV
Abstract:	The authors conducted a scientific study of the actual problem of assessing the economic efficiency of investment projects at high-tech industrial enterprises. In the process of organizational and economic analysis, the reconstruction of industrial enterprises of the aerospace complex and the most effective methods for justifying financial investments in the process of reproduction of all fixed assets were studied. A detailed justification has been carried out for the optimal choice of the most attractive investment projects for aviation and rocket and space enterprises, which are recommended to be included in the state investment program for the digitalization of the national economy. The author's classification of new criteria is proposed, depending on the numerous requests of all participants in investment construction in the process of reconstruction and technical re-equipment of industrial enterprises. The authors assign a special role to a detailed analysis of investment projects to reform advanced industrial enterprises that are actively introducing modern digital technologies in their work.
Keywords:	Digital Economy, Digital Transformation, Enterprises Of Rocket And Space Industry, Human Resource Management.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	DEVELOPMENT OF MACHINE LEARNING BASED EMPIRICAL MODEL FOR ESTIMATION OF SOLAR RADIATION
Author:	SAUMYA MISHRA, DEEPENDRA PANDEY, SAURABH BHARDWAJ, AMAR CHOUDHARY
Abstract:	This paper proposes Optimal harvesting of solar radiation is a thrust area since last decades. Machine Learning Techniques are more advocating technology for estimation related cases. This study claims that extreme learning machines (ELM) can be able to estimate solar radiation more accurately than any other system. The proposed technique was evaluated based on data collected in Karaman Province between 2010 and 2018. This evaluation was carried out with MATLAB 2022 under toolbox- Multiprocessing Algebra Data (Mu PAD). It was hypothesized that ELM had shown a better estimating performance than the other method when the data were compared. In addition, ELM was evaluated using a variety of activation functions to find the one that provided the most accurate estimate response. ELM models have minimum RMSE which is 14.02 as compared to other methods. The computed data and the approximation data that are paired together to form the global model have a superior correlation coefficient, which is about 95.67% on standard. This signifies the global model is accurate to a high degree. A comparison between ELM and other training algorithms reveals that it is the most effective strategy for training all sub models based on the data collected in the two rounds of the study. All ELM activation functions performed better than all transfer functions of SCG and RP when compared to the general outcomes of the evaluations. Comparing the accuracy and reliability of the suggested machine-learning algorithm to the current industry standard in predicting solar radiation time series for solar energy system design and management.
Keywords:	Extreme Learning Machines, Solar Radiation, Machine Learning, Artificial Neural Network, Root Mean Square Error, Mean Absolute Percentage Error.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	MODEL OF PROTECTING DATA IN THE CLOUD
Author:	OSAMA ALJUMAIAH, MOUNIR FRIKHA
Abstract:	With the wide spread of cloud services, people use cloud technology to store their personal information. Technologies have a significant part in the operation of our lives. The most major issue is data breaches caused by hackers. Therefore, we must provide a cryptographic technique that cannot be broken to safeguard the data from hackers. Cryptography encrypts data so that only the authorized user may decode it. This project aims to propose a protection model for data in the cloud using RSA Algorithms.
Keywords:	Cloud, Security, Privacy, Encryption, Data
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	IMPROVING LOW-LIGHT FACE EXPRESSION RECOGNITION USING COMBINATION OF MIRNET AND RESNET50 MODELS
Author:	RHEIVANT BOSCO THEOFFILUS, GEDE PUTRA KUSUMA
Abstract:	Emotion is the part of humans expressing their feeling about something. There are seven types of emotions in humans which are anger, disgust, fear, happiness, neutral, sad, and surprise. These human expressions can be examined and exploited in research fields, say, psychologists. Numerous areas, including the economic and health sectors, could benefit from this research. For instance, recognizing user expressions while a player is playing a video game in dark conditions. Other applications of the health industry include spotting furious driving expressions in cars at night. However, this topic still has certain shortcomings, one of which is illumination, which is unfortunate. Therefore, in this work, we propose a combination of low-light image enhancement and face expression recognition (FER) for recognizing expression in low-light conditions. MIRNet, RetinexNet, Retinex, and AGC are the four models that were used in this study for image enhancement. While the FER model consists of two models, ResNet50 and Inception ResNet V2. The result of the best combination of FER and image enhancement is MIRNet scratch and ResNet50 with 67.6% accuracy in low-light conditions. In our experiment, this combination of FER and image enhancement has the best accuracy.
Keywords:	Face Expression Recognition, Low Light Image, Image Enhancement, ResNet50 Model, Deep Learning.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	A SYSTEMATIC LITERATURE REVIEW ON OPEN GOVERNMENT DATA: CHALLENGES AND MAPPED SOLUTIONS
Author:	WAHYU SETIAWAN WIBOWO, DANA INDRA SENSUSE, SOFIAN LUSA, PRASETYO ADI WIBOWO PUTRO, ALIVIA YULFITRI
Abstract:	Open Government Data (OGD) has grown exponentially in the last few years and has served as the bedrock of a data-driven nation. OGD helps government promotes transparency and foster innovation throughout the nation. As Indonesia promulgated Satu Data Indonesia (SDI) policy in 2019, the country faces challenges in implementing SDI. Using a systematic review method, Preferred Reporting Items for Systematic Review and Meta-analyses Method (PRISMA) 2020, this research aims to uncover the challenges of implementing OGD and map the possible solutions to address the barriers. The authors discover 23 challenges in OGD implementation grouped by the TOE framework and 12 remedies grouped by the UTAUT framework. Organisational barriers become the most common problem in OGD initiatives, while solutions in facilitating conditions constitute the most common solution. The authors then mapped the challenges into the solutions. This study therefore could assist other researchers in OGD-related studies and provide governments with an in-depth reference for OGD implementation guidelines in the future.
Keywords:	Open Data, Open Government Data, OGD, Government, Systematic Literature Review, Challenges, Solutions, TOE Framework, UTAUT Framework
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	CONTINUANCE INTENTION TO SUBSCRIBE TO A VIDEO-ON-DEMAND SERVICE: A STUDY OF NETFLIX USERS IN INDONESIA
Author:	DANIEL YEDIJAOKTO SULAIMAN, VIANY UTAMI TJHIN
Abstract:	With the growing interest in the media and entertainment industry, the competition between players is becoming even tighter, thus having customers attached to a particular subscription video-on-demand (SVOD) brand is important to ensure subscription continuance. Despite the growing research in the SVOD services, no study has investigated the influence of inertia as an attachment factor and satisfaction on the users’ continuance intention on SVOD particularly. Therefore, this study aims to address the lack of literature on the continuance model by testing the influence of content quality, system quality, trust in mobile payment, and price fairness, and further test the impact of inertia and satisfaction towards one’s continuance intention to subscribe to a particular SVOD brand. 532 individuals partake in an online survey to give their perspective on their current SVOD subscription which is then analyzed with partial least square-structural equation modelling (PLS-SEM). The results indicate that trust in mobile payment and content quality are the direct determinants of continuance intention to subscribe, and that user satisfaction and inertia successfully mediate all the independent constructs used in this study towards continuance intention to subscribe. The findings of this study is useful for SVOD players in understanding the importance of each factor in order to strategize and fulfill their customer demands to assure long-term viability of their businesses.
Keywords:	Content Quality, System Quality, Trust in Mobile Payment, Price Fairness, Inertia, Continuance Intention to Subscribe
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	MRI BASED CONVOLUTIONAL NEURAL NETWORK MODEL WITH ASSOCIATED FEATURE VECTOR MODEL FOR GRADE DETECTION OF MULTIPLE SCLEROSIS LESIONS
Author:	KALPANA NARAYAN RODE, RAJASHEKAR JANGAM SIDDAMALLAIAH
Abstract:	Multiple Sclerosis (MS) patients who are receiving treatment must be monitored for new or expanded white-matter lesions. Multiple Sclerosis patients' multimodal brain Magnetic Resonance Imaging (MRI) images have been used to develop the first fully automated end-to-end deep learning approach for the prediction of future disability progression detection. Convolutional neural networks are used in nearly all new segmentation methods, each with a unique architecture. MRI scans are used to detect lesions in the brain and spinal cord that are characteristic of MS, and these scans are also used to track the disease's progress over time. The detection of lesions by hand takes a long time and is prone to error. It is difficult to detect the lesions manually, especially in the grey matter. Convolutional Neural Networks (CNNs) can be used to automatically segment lesions. In this research, a CNN model is used for the detection of Multiple Sclerosis. The use of CNN to extract features from flare MRI improves recognition performance. First, a segmentation-accurate CNN network is implemented, and then a false-positive reduction network is added to increase efficiency even further. CNN replaces the fully connected layer at the end of two parallel convolutional pathways, which are concatenated together. In this research, an Associated Feature Vector Model for Grade Detection of Multiple Sclerosis Lesions (AFVM-GD-MSL) using CNN is proposed. In terms of performance evaluation, CNN is a good choice because of its lack of noise sensitivity. No lesion segmentation was necessary to achieve the high detection rates achieved by CNN. As long as the variability is taken into account, this fully automated method yields accurate MRI analysis and grade classification.
Keywords:	Multiple Sclerosis, Convolutional Neural Networks, Segmentation, Feature Extraction, Grade Classification.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	MULTI LINKED RELEVANT FEATURE SET MARKING MODEL USING CONVOLUTION NEURAL NETWORK FOR DIABETIC RETINOPATHY GRADE DETECTION
Author:	LUBNA TARANUM MAZAR PASHA, JANGAM SIDDAMALLAIAH RAJASHEKAR
Abstract:	Diabetic Retinopathy (DR) is the most common cause of blindness in 12% of the population every year. One of the most effective ways to prevent vision loss and reduce health care costs is to screen for DR and monitor disease progression in the early stages, especially in the asymptomatic stages. Nonmydriatic digital colour fundus cameras are commonly used to take colour images of the retina in most screening programmes. Early diagnosis and treatment can significantly reduce the risk of severe vision loss. DR screening has been found to be a cost-effective method of reducing the burden on the health care system. The development of automated tools to aid in the detection and evaluation of DR lesions that has been a major focus of recent research. Patients who are diagnosed with retinopathy early on have a better chance of preserving their vision. The diabetic retinopathy grade severity level assessment is proposed in this paper that uses Multi Linked Relevant Feature SetMarking Model using Convolution Neural Network (MLRFSMM-CNN). The standard DIARECTDB1 datasets are used to obtain the colour retina images. Images of the retina are first analysed to identify lesions, such as blood vessels and haemorrhages, as well as exudate and microaneurysms, on the retina. This is followed by extraction of various relevant features, such as area of the segmented exudates, the quantity of microaneurysms in the segmented image, the mean and standard deviation of segmented lesions, to determine the grade severity level of the disease. The proposed model is compared with the traditional methods and the results represent that the proposed model performance is enhanced.
Keywords:	Diabetic Retinopathy, Feature Set, Feature Extraction, Convolution Neural Network, Grade Detection.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	A COMPARATIVE STUDY USING IMPROVED LSTM /GRU FOR HUMAN ACTION RECOGNITION
Author:	AZHEE WRIA MUHAMAD, AREE ALI MOHAMMED
Abstract:	A key focus of this paper is the automatic recognition of human actions in videos. Human action recognition is the automatic understanding of what actions occur in a video implement by a human. Deep learning models can be used to identify and classify objects accurately. The process of detecting human actions in videos is known as human action recognition. One of the deep learning algorithms for sequence data analysis is a recurrent neural network (RNN). In a conventional neural network, the inputs and the outputs are independent of each other. At the same time, RNN is considered a type of Neural Network where the output from the previous step feeds information to the current phase. It has many applications, including video sentiment classification, speech tagging, and machine translation. Recurrent networks are also distributed parameters across each layer of the network. Several layers are stacked together to increase depth in forwarding and backward information of long short-term memory (LSTM) and Gated Recurrent Unit (GRU). This paper proposes two models for various action recognitions using LSTM and GRU, respectively. The first model was improved by increasing the LSTM layers to four and the number of units in each layer to 128 cells. While in the second model, GRU layers were extended to two layers with 128 cells, and the (update and reset) gates are modified based on the previous and current input. A comparative study was conducted during the experimental tests performed on the UCF101 action dataset regarding the accuracy rate for both models. Test results indicate that the accuracy has a significant improvement compared with other state-of-the-arts action recognitions, which are 95.19 % and 92.9 % for both improved LSTM and GRU, respectively.
Keywords:	Action Recognition, Deep learning, LSTM/GRU, Performance Accuracy, and RNN.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	METHODOLOGY FOR CREATING AND CONDUCTING RESEARCH ON DIGITAL TWINS OF TRANSFORMERS AT THE STAGE OF RESEARCH WORK
Author:	VLADIMIR STROEV, ANDREY SAZONOV , ALEXEY TIKHONOV , DIEGO FELIPE ARBELÁEZ CAMPILLO
Abstract:	The article is devoted to the problems of creating simulation models of products (transformers) based on the concept of digital twins, since due to the rather large spread in the properties of electrical materials, models of this level need to be corrected for the information received from a really working device. The authors propose a methodology for the development of digital twins for transformers, which will ensure cost reduction, reduce the time spent on their design. The theoretical and methodological base of the research is based on the scientific works of Russian specialists, as well as on the use by the authors of various numerical methods for modeling physical circuits in the Matlab environment. The authors propose a methodology for the development and application of the concept of digital twins of transformers, which will optimize and improve the process of developing and making design decisions.
Keywords:	DDigital Twin, Intelligent Monitoring System, Power Transformers, Simulation, Dynamic Models, Life Cycle
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	METHODS OF COMPUTER MODELING, APPROXIMATION AND MECHANICAL BEHAVIOR OF ROBOTIC SYSTEMS UNDER DYNAMIC INFLUENCES
Author:	KAMIL KHAYRNASOV, ANTON SOKOLSKIY, VLADIMIR ISAEV, DANIL SAPRONOV
Abstract:	The methodology of modeling and analysis of the stress-strain state of a three-stage dynamic semi-natural modeling stand (stands) made of homogeneous material under dynamic loads is considered. Semi-simulation stands are used to simulate the flight characteristics of aircraft in ground conditions. The simulators are a complex structure consisting of moving elements connected by bearings, gearboxes, gears and engines. Accounting for the interaction of these elements is a complex task that requires modeling that is identical to the behavior of the stand elements in real conditions. One of the main parameters characterizing the effective workability of a test stand is its inertia characteristics, which allow to quickly and accurately implementing the given commands. One of the most effective and tested methods of structural studies is the finite element method. In the present study the methodology of modeling bearings, reducers, gear wheels in the stand by replacing them with systems identical in stiffness and deformability taking into account the applied research method, the method of finite elements. As a result of the study the methods of modeling a complex dynamic structure containing bearings, reducers and gearboxes. Methodology for modeling of three-layer structures consists of external bearing layers and a filler layer between them, preventing shear stresses and convergence of bearing layers. The stress-strain state of the stand under dynamic loads is investigated.
Keywords:	Method, Modeling, Approximation, Stands, Finite element method, Mechanical behavior, Dynamic loading.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	AN EFFICIENT TWO-FACTOR USER AUTHENTICATION PROTOCOL FOR AD-HOC WIRELESS SENSOR NETWORKS
Author:	V.S.SUDHAKARA RAO ANDE, SREENIVASULU MERUVA
Abstract:	Providing access to confidential messages in a secured manner within Ad-hoc WSNs (Wireless Sensor Networks) is the challenging issue for researchers, due to lack of physical security and a greater number of potential attacks on the information transmitted through wireless radio. In 2021, Tsu-Yang et. al, presented a two-factor authentication protocol for Ad-hoc WSNs with the usage of smart card. It is an efficient scheme. This reduces the sensor node's energy usage while performing authentication of a user. It suffers from off-line password computation attack, the user un-traceability attack, password recovery attack. We realized that Tsu-Yang et. al's, scheme failed in real-time Ad-hoc WSN, where the information can be delivered in rigid time constraints. It also increases the burden on Gate Way Node (GWN) and leads to a denial-of-service attack. So, we present an authentication scheme that would be both effective and reliable, for Ad-hoc WSN to deliver information in a secured manner and in rigid time constraints. The security level of the protocol to be proposed is evaluated by the usage of Automated Validation of Internet Security Protocols and Applications (AVISPA) tool.
Keywords:	Ad-hoc WSN; Authentication Protocol; Gate Way Node, Rigid Time Constraint; Smart Card.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	HEALTH SOCIAL CREDIT SYSTEM SIMULATION OF COVID-19 PANDEMIC IN TANGERANG CITY DISTRICTS
Author:	YUSUF SUDIYONO, AGUNG TRISETYARSO, MEYLIANA, HARJANTO PRABOWO
Abstract:	The spread of Covid-19 massively occurred in 2019 resulted in significant rate of morbidity and mortality. This study aimed to analyze data of factors contributing in the rise of Covid-19 cases. To analyze such factors, a comprehensive and thorough analysis of Covid-19 spread data in Tangerang city is required. Data evaluation have been conducted as of the occurrence of Covid-19 in early of 2020 until September 2022. From that data, an analysis of Covid-19 spread was conducted based on case phase and level. In the case, phase-based analysis, accumulation in each case number of each semester as well as analysis of factors that may bring any effect on the increase of Covid-19 cases have been conducted. While in case level, analysis of Covid-19 spread per semester has been done to measure the level of severity of its potential impacts. Such severity level was discovered from the total confirmed cases, the number of recovered cases, and the mortality rate arising in each level. The perilous level stage occured at level 4 in which the cases and its mortality cases are considerably high. Hence, in the area with the highest case, strict control and surveillance functions can be executed during the outbreak of Covid-19. Social credit system may be done by providing an evaluation and appreciation to regions with the lowest cases for Covid-19 spread. Dataset used in this paper was the data of Covid-19 spread in Tangerang city.
Keywords:	Covid-19, Case Levels, Case Phases, Integration Data, Tangerang City
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	HYBRID ENCODING AND DATA TRANSFORMATION FOR CLOUD DATA SECURITY IN POST-QUANTUM ERA
Author:	SHAIK MOHAMMAD ILIAS, V.CERONMANI SHARMILA
Abstract:	In the wake of emergence of quantum computers, there have been efforts on Post Quantum Cryptography (PQC) schemes. PQC schemes can replace the classical cryptographic primitives for stronger level of security. Demand for such schemes is driven by initiatives and interests from security standardization bodies across the world. In response to the demand, many schemes came into existence. Motivated by the PQC requirements, in this paper, we proposed a security algorithm known as Hybrid Encoding and Data Transformation (HEDT). It has provision for data owners to have more secure outsourcing of their data to cloud and retrieval of the same. It has multiple data transformations with data encoding and decoding to have stronger level of security. It not only provides security to outsourced data but also fulfils needs of features such as data integrity and data availability. It is designed to be a candidate for PQC requirements to safeguard not only data at rest in cloud but also data in transit. An empirical study revealed that the HEDT is capable of providing stronger level of security to data outsourced to cloud.
Keywords:	Encryption, Decryption, Encoding, Decoding, Cloud Data Security, Cloud Computing
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	USING STRING SIMILARITY ALGORITHMS TO FIND ENGLISH WORDS POTENTIALLY ORIGINATING FROM ARABIC
Author:	MAJED ABUSAFIYA
Abstract:	One interesting linguistic issue is the study of English words originating from Arabic. This is based on identifying pairs of Arabic and English words that show obvious similarity in pronunciation and meaning. To the best of the knowledge of the author, no computational solution was proposed to support this subject. In this paper, similarity algorithms are used to measure the similarity between Arabic roots and English words. The proposed solution is implemented and the resulting similarity data was studied. One contribution of this work is finding many pairs of Arabic and English words that did not exist in the widest collective reference dictionary that is known in this context. Another contribution is showing how the string similarity algorithms may be utilized to this purely linguistic issue.
Keywords:	Algorithms, String Similarity, Computational Linguistics, Arabic, English
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	EVALUATION OF DEEP LEARNING AND MACHINE LEARNING ALGORITHMS IN INTRUSION DETECTION SYSTEMS
Author:	SAWSAN ALSHATTNAWI
Abstract:	The advances in using technologies and the Internet increase the spread of data over the network. New attacks may be generated to threaten network security and data integrity, in addition, intruders always exist and we can't ignore their presence. Many tools and algorithms have been used to create many intrusion detection systems. A Network Intrusion Detection System (NIDS) is one of these tools designed to detect intrusion and stop it. Machine Learning (ML) and Deep Learning (DL) algorithms are used to build powerful NIDS. Deep learning proved its success by giving high accuracy and best performance compared with machine learning algorithms. In this paper, we will compare the accuracy of the used approaches by reviewing the literature and focusing on the research that used a famous and well-known dataset called KDD Cup’99. The paper will, in the end, summarize the issues and challenges in NIDS. The study summarizes the main issues and challenges of using the machine and deep learning and advice the researchers to start new approaches such as the deep forest over more recent datasets.
Keywords:	Network Intrusion Detection systems, Machine Learning, Deep Learning, KDD Cup 99.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	AN EXPERIMENTAL INVESTIGATION BASED ON SERVICES OF VIDEO STREAMING USING DEEP NEURAL NETWORK FOR CONTINUOUS QOE PREDICTION
Author:	SK. MAHABOOB BASHA, S.A. KALAISELVAN
Abstract:	Now-a-days Global Internet traffic is created by video streaming which is a primary source of platform, this may create and contact with the worldwide audience. The user the contents high quality streaming to delivering a crucial role play with quality of experiences the continuous user predicting with video streaming services. By the temporal dependencies that can cause the complexity in data QoE and the factor influence QoE among non-linear relationship that can introduced challenge to prediction QoE continuous. To effectively capture Gated Recurrent Unit (GRU) that utilized the existing studies this can be deal, accuracy QoE prediction excellent resulting. Even-through, GRU complexity is high computational, architecture with characteristic processing sequential with it, power computational with limited devices it’s performance about the serious question which has been raised. Meanwhile, Deep neural network with variation of Temporal convolutional network (TCN), for modelling task with sequences which has been proposed, computational complexity and accuracy prediction with terms GRU based on the method of baseline over the performances prediction supervisor has been provided. In this paper, based on the model TCN with improved, namely DNN-QOE, the QoE proposed is predictive continuous, sequential data with characteristic pose, to overcome the complexity computation based on the advantage of TCN leverage from drawback model of GRU based QoE, to improve the accuracy of the QoE prediction they improved the architecture that has been introduced at the certain time. The performances of the DNN-QoE are highly competitive.
Keywords:	Services, Video Streaming , Neural Net, QOE Prediction
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	PREDICTION OF AUTO-DETECTION FOR TRACKING OF SUB-NANO SCALE PARTICLE IN 2D AND 3D USING SVM-BASED DEEP LEARNING
Author:	DR.P.S.RAMESH, DR.I.SUDHA, DR.N.SATHEESH, DR.G.GOVINDA RAJULU, S.A.KALAISELVAN
Abstract:	The time series position with large video files that requires conversion based powerful biophysical tools with particle tracking i.e., data analysis the species on interest with traces. The particle tracking that obtained accuracy that demonstrate with Nano scale tracking particle based on the method deep learning. The method current tracking, bright object identify based on input parameter with set limited, heterogeneity spatiotemporal spectrum that handle ill-equipped and biological environment complex with submicron species can present typical poor ratio signal-to-noise. Method execute tracking and optimize which frequently necessary with the involvement of user extensive, the user bias was introduces,which was not inefficient. Method to develop automated tracking, data image from localization particle based on support vector machine that has been developed, method tracking fully automated, 6000 parameter comprising, and video conditions diverse portfolio to train the network that is used with technique deep learning. Accuracy and automation unprecedented has been provided with track of support vector machine, 2D and 3D video simulated with rates false negative and positive low exceptionally and species difficult-to-track video 2D experimental video.
Keywords:	Particle Tracking, Deep Learning, Support Vector Machine, Bio-Imaging, Quantitative Biology.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	FORECASTING OF ORIGIN-TO-DESTINATION REQUESTS FOR TAXIS USING DNN ALGORITHM WITH NYU DATABASE
Author:	R.B.SAROORAJ, S. PRAYLA SHYRY
Abstract:	Prediction of taxi requests haslately drawn increased much attention in research owing to its high applicability in massive intelligent transport systems. Most previous methods, on the other hand, focused solely on predicting taxi demand in origin areas, ignoring the analysis of target passengers' special circumstances.We believe it is inefficient to assign cabs to all areas in advance solely based on taxi origin request. This work studies a critical and fascinating task known as taxi origin-destination demand prediction, whose purpose is to forecast future cab requests across pairs in all areas. Determining the process to collect contextual data effectively in order to learn demand patterns. A novel methodology known as the Deep Neural Network (DNN) with Deep learning-based models is focused in this paper, which outperforms traditional machine learning methods in a variety of classification tasks, including origin and destination views. Extensive tests and analyses on a broad dataset clearly show that our DNN outperforms several other methodologies from literature.
Keywords:	Taxi Origin-Destination, Spatial-Temporal Modeling, Deep Neural Network, DNN Algorithm, NYU Database.
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	SPATIAL TRANSFORMER NETWORKS FOR ADAPTIVE VIDEO ENCODING AND DECODING IN VIDEO TRANSMISSION
Author:	AJITHA G, SANTHIPRABHA I, AJITHA G
Abstract:	Video encoding is one of the methods for improving video transmission quality. Recently adaptive and scalable video encoding technology is emerging which compresses the video files while maintaining their quality. However, existing methods of video encoding and decoding are used to extract the contextual information, which does not suited for all kinds of devices, applications and services in wireless networks. Hence, the new video encoding and decoding technique is presented for video quality improvement using deep learning. For that purpose, Spatial Transformer Networks (STN) is presented in this paper. Firstly, coefficients are extracted in the set of frames of a given input video. This is done by means of fast curvelet transform method that splits the frame into multiple levels. Secondly, approximate coefficients generation is implemented by means of spatial transformer networks. Then the detailed coefficients are optimized by means of the Human Mental Search Optimization technique. Later, the side information is extracted using Inverse Curvelt Transform. Finally the, video quality is evaluated using Fuzzy Logic Inference System (FLIS) using video quality assessment metrics. This feedback is forwarded to the source node. The comparison is between the proposed approach with the existing approaches in terms of power consumption, time, PSNR, MSE and RMSE for different datasets.
Keywords:	Video Encoding And Decoding, Spatial Transformer Networks, Fuzzy Logic, Side Information Extraction, Video Quality Assessment
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	HUMAN FACTOR RELIABILITY IN SAFETY CRITICAL SYSTEMS: A GLOBAL APPROACH FOR SPECIFIC CONTEXTS
Author:	JIHAD OUAHLI, ABDELGHANI CHERKAOUI
Abstract:	The human reliability evaluation and the risk management of human factor activity have been the subject of several research works for half a century. Many approaches for evaluating human reliability have been proposed as well as predictive cognitive models of operator performance. However, we still observe incidents and accidents on a daily basis that mainly blame the failure of the human factor. In this article, we review the state of the art in the field of reliability of the human factor and the main contributions in that field. We then propose a holistic and operational approach for a global evaluation of the performance of the human factor reliability.
Keywords:	Human Factor Reliability, Safety Critical Systems, Human Error, Human And Organizational Factors
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	A RELAY-BASED AUTOMATIC BALANCING SYSTEM FOR THREE-PHASE LOADS IN INDUSTRY 4.0
Author:	YOUSSEF ZERGUIT, YOUNES HAMMOUDI, IDRISS IDRISSI, MOSTAFA DERRHI
Abstract:	The challenges of the existing industry to remain competitive have prompted the exploration of novel techniques that increase production, lessen waste, and enhance efficiency. The term "Industry 4.0" describes a new way of thinking about how technology might be used in manufacturing, mainly related to the Internet of Things (IoT), Big Data, and Cyber-Physical Systems (CPS). One of the main objectives of this fourth industrial revolution is the complete digitalization of the manufacturing chain in order to realize the interconnection of people, machines and devices. In light of this revolution, the investigation of an automatic balancing system (ABS) is something that interests us. In fact, an ABS is utilized to monitor and control three-phase loads in real time. The objective of this paper is to propose and implement a relay-based automatic balancing system using single-board computers (SBCs) as the main element to control the system operations, where sensors measure parameters like current, voltage, power, etc., in a three-phase load. The proposed system is capable of balancing the three-phase load by automatically switching the load phases between them, thus avoiding unbalanced load operation and the resulting problems.
Keywords:	Automatic Balancing System (ABS), Three-Phase Loads, Three-Phase Balancing, Industry 4.0, Industrial Internet Of Things (IIOT).
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text

Title:	DAE-BILSTM: A FOG-BASED INTRUSION DETECTION MODEL USING DEEP LEARNING FOR IOT
Author:	IBRAHIM MOHSEN SELIM, ROWAYDA A. SADEK
Abstract:	Fog computing efficiently brings services to the edge network because it facilitates processing, communication, and storage to be closer to the edge devices. Fog computing is principally used for the Internet of Things (IoT) instead of cloud computing as an ideal option to reduce latency, especially for time-sensitive applications. Many different security challenges and cyber-attacks have emerged in the fog layer, rendering typical defense solutions ineffective. The Network Intrusion Detection System (NIDS) is a security system that uses various advanced learning approaches to detect and predict various attacks. IoT devices lack the computing power and energy needed for such, hence, IDS must be at the fog node to monitor data coming from different sources, such as IoT devices or neighboring fog nodes. To identify cyber-attacks and malicious states in IoT network traffic, this paper proposes an intrusion detection model named DAE-BiLSTM based on deep learning (DL) in the fog layer. The DAE-BiLSTM model employs a Deep AutoEncoder (DAE) in conjunction with Bidirectional Long Short-Term Memory (BiLSTM). Many existing state-of-the-art studies used the NSL-KDD dataset, which is now out of date and does not include new IoT cyber-attacks. The results of these studies also need to be improved by using advanced DL techniques to be effective in detecting current cyber-attacks. So the network-based ToN_IoT dataset is used for the training and testing of the DAE-BiLSTM model. The ToN_IoT dataset comprises new and better-suited cyber-attacks for IoT. The DAE-BiLSTM model obtained a high accuracy rate of 99.7%, 99.4% for precision, 99.7% for recall, and 99.5% for the F1-score in classifying normal traffic data from attacked traffic data in the fog layer.
Keywords:	Internet of Things (IoT), Fog Computing, Intrusion Detection System (IDS), ToN_IoT dataset, Deep Learning (DL)
Source:	Journal of Theoretical and Applied Information Technology 15^th MArch 2023 -- Vol. 101. No. 5-- 2023
Full Text