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

Welcome to Journal of Theoretical and Applied Information Technology

Submit Paper / Call for Papers

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

Journal of Theoretical and Applied Information Technology

May 2023 | Vol. 101 No.10

Title:	MENTAL HEALTH ANALYSIS USING NATURAL LANGUAGE PROCESSING
Author:	K.DEEPA, C.RANJEETH KUMAR, M. KALEEL RAHMAN, E. DERRICK GILCHRIST
Abstract:	Many people all over the world are depressed and are completely unaware of it. Depression is a mental illness in which a person is constantly unhappy and loses interest in almost everything. Depression can result in self-harm or even suicide. People can, thankfully, recover from depression with the help of therapy and medication. When a person's depression is detected early, his or her recovery will be greatly aided. Our project's main goal is to detect the depression in users' speech while also providing assistance for depression recovery. Nlp models such as word embedding and tone analyzer are used to detect depression, and recovery guidance given to the patient by providing the consultant details in their surroundings.
Keywords:	CNN, Word Embedding. Tonal Analysis, Depression Detection.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	INNOVATIVE METHOD OF CLASSIFICATION OF PULMONARY NODULES USING 3D CNN ARCHITECTURE
Author:	KISHORE SEBASTIAN, S. DEVI
Abstract:	Pulmonary nodules are spots or lesions that are diagnosed in the lungs and can have both benign and malignant causes, mainly related to lung cancer. According to statistics, lung cancer is in seventh place and is the most lethal. Various international medical institutions are working to improve the diagnosis of lung cancer, since the main cause of death is late diagnosis. For this reason, the analysis of pulmonary nodules is a challenge in the processing of medical images to determine the appropriate treatment, new methodologies and techniques are proposed not only by experts in medicine but also in by other scientists. In this work 3D Convolutional Neural Network is used to classify pulmonary nodules in CT images. The proposed architecture results, with an accuracy of 0.9076, kappa of 0.7773, sensitivity of 0.8483, and specificity of 0.9321 and AUC of 0.8903, in the classification of pulmonary nodules.
Keywords:	Lung Cancer, Lung Nodule Classification, Machine Learning
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	LAXITY-AWARE MIXED-CRITICALITYTASK SCHEDULING FOR ENERGY-EFFICIENT HETEROGENEOUS MULTICORE PROCESSORS
Author:	N.GOMATHI, K.NAGALAKSHMI
Abstract:	Mixed-criticality systems (MCS) have developed as an efficient solution in several industries, where numerous tasks with different criticality levels (safety requirements) are assimilated onto a shared computational platform. Today, increased energy consumption in MCS, especially in critical situations, leads to temperature hotspots, which may disrupt the reliability and correctness of the system. As processors with multiple processing elements are becoming the vital paradigm in MCS, an integrated timeliness and power management is an important issue. This paper proposes a laxity-aware mixed-critical task scheduling (LMTS) algorithm that provides correctness, timeliness, power management, and guaranteed service level in MCS simutaneously. This method minimizes energy consumption of the system considerably through dynamic voltage and frequency scaling (DVFS) method. It collects several workloads concurrently and form clusters with one high-critical workload and a set of low-critical workloads. It determines the laxities and selects the most suitable cluster to exploit the available laxity based on its impact on the energy consumption and hotspot problems of the system. However, changing the core frequency, allocating more suitable cluster for available laxity, and finding out an appropriate core for mapping at runtime are difficult processes and cause deadline desecration which is not suitable for safety-critical tasks. Therefore, we develop an effective scheduling method using DVFS schemes and task migration techniques in online mode to utilize available laxity. We also defined cost functions to select the most apposite cluster to right core by scaling its voltage/frequency (v/f) value or to migrate it to another processing element. We assess the effectiveness of our scheduling algorithm in a heterogeneous multicore processor with real-time tasks.
Keywords:	Arm Big. LITTLE; DVFS; Energy Efficiency; Task Scheduling; Mixed-Criticality System; Multi-Core Processors; Laxity Utilization.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	A MODIFIED SVM ALGORITHM TO ENHANCE THE CANCER CLASSIFICATION
Author:	RETHINA KUMAR, GOPINATH GANAPATHY, JEONG-JIN KANG
Abstract:	The survival rate of breast cancer patients has increased due to the advancements in the treatment of the disease. These include the use of newer and more effective drugs. There are various types of breast cancer, which can be treated through different methods. Currently, there are numerous studies that are focused on developing a better understanding of this disease. In order to improve the classification of breast cancer, a new machine learning algorithm is proposed. This method uses support vector machine learning to enhance the performance of existing models. The proposed model can rectify the inconsistencies in the existing breast cancer dataset and improve its performance. It can also create a high-quality Wisconsin Diagnostic Breast Cancer (WDBC) data set. The proposed model can then predict the likelihood of a patient developing breast cancer. It can also diagnose the patient based on the data collected. The researchers were able to test the proposed model against several machine learning models. They were able to achieve high accuracy levels.
Keywords:	Breast Cancer, Machine Learning, Diagnosis, Prediction, Benign, Malignant.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	VARIATION OF DIGITAL MARKETING STRATEGY ON TOURISM BUSINESS SCALE IN BALI, INDONESIA
Author:	NOELLA LUDYLANE TERSIANA NENDISSA, AGUIRA FORTUNA, SISWANTINI
Abstract:	With all its beauty, Bali can attract various people from all over. With more and more visitors to Bali, there are opportunities for business people with various types of businesses. With so many businesses, it must pay attention to every critical point in doing business. One of the critical points that a business must consider is a marketing communication strategy. A marketing strategy is needed to survive among other businesses, especially digital marketing communication, considering the times that continue to prioritize digital. This study aims to determine whether differences in the size of the business scale can affect marketing strategies carried out in businesses in Bali. The research method used is qualitative with a descriptive approach. This research was researched the numbers of tourism services ranging from clothing brands to hotel services. The result of the article is the variances in the size of a company's scale have an effect on the marketing technique it employs; micro-business to large businesses have different types of digital marketing, but there are the marketing strategies of these four business scales that similar comparable since they both employ social media as a marketing tool.
Keywords:	Bali, Marketing Strategy, Digital Marketing Communication, Business, Tourism Marketing
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	NETWORK INTRUSION DETECTION BASED ON ONE-DIMENSIONAL CNN WITH CHIMP OPTIMIZATION ALGORITHM
Author:	Dr. V. GOKULA KRISHNAN, Dr. M. V. VIJAYA SARADHI, Dr. S. VENKATA LAKSHMI, S. KAVIARASAN4, ABOTHU GEETHA
Abstract:	The widespread development of the Internet of Things (IoT) is known throughout the world. The 2016 Dyn cyberattack revealed significant flaws among smart grids. IoT security has become one of the top concerns. The security of the entire IoT environment is affected by the affected networks that are connected to the risks posed by contacts. Nowadays, the diversity and complexity are evolved by defense attack vectors in recent times. It is one of the important things to prevent, identify or detect the new attacks in the IoT environment by analyzing the techniques. Therefore, network intrusion detection systems (NIDS) play an important role in protecting computer networks. Detection of security-related events using machine learning approaches has been extensively explored in the past. In particular, machine learning-based web browsing detection has attracted a lot of attention due to its ability to detect unknown attacks. Many classification techniques such as Decision tree (DT), Support Vector Machine (SVM) have been used for that purpose, but they were mostly classical schemes, like final trees. In this study, the use of deep learning technique is explored for NIDS. Initially, the noise samples are minimized in the majority segment by using One-Sided Selection (OSS) and then, Synthetic Minority Over-sampling Technique (SMOTE) is used to develop the minority samples for creating the balanced datasets. In this way, the research work is used to fully understand the characteristics of minority models and greatly reduce the sample training time. Second, we use a one-dimensional convolutional neural network (1D-CNN) with the Chimp optimization algorithm (COA) to extract the features, creating a hierarchical network model (HNM). The research work tested the classification accuracy of CNN-COA with existing techniques and its performance is verified by experiments in the NSL-KDD database. The proposed model achieved 87.19% of accuracy, 88% to 89% of precision and recall, where the existing model CNN achieved 81.75% of accuracy and 82% of precision and recall.
Keywords:	Attacks, Chimp Optimization Algorithm, Network Intrusion Detection Systems, One-Side Selection, Synthetic Minority Over-sampling Technique, UNSW-NB15.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	SINGLE BOOTSTRAP APPROACH WITH GEOGRAPHICALLY WEIGHT REGRESSION MODELING USING PARTICLE-SIZE FRACTION
Author:	HENNY PRAMOEDYO, WIGBERTUS NGABU, SATIVANDI RIZA, LUTHFATUL AMALIANA
Abstract:	The challenge of the high need for soil spatial data information has led to the rapid development of spatial modeling for soil attributes in the last few decades. Soil texture is an essential attribute that determines the direction of soil management and must be modeled accurately. However, on the other hand, soil texture is a soil attribute that is relatively difficult to model because it is a compositional data set. The difficulty that arises from this compositional data set is the limitation of constant quantities; namely, the sum of the fractions of sand, silt, and clay must be 100%. Through DEM data, topographical variability can be obtained so that it will be a predictor or independent variable in predicting soil texture. In addition, Geographically Weighted Regression (GWR) was also used in this study to pay attention to the effect of spatial heterogeneity. It uses the bootstrap method with the GWR model to overcome bias in the model parameters. Residual bootstrap is a bootstrap method that is applied to the residual resampling process. The aims of this study: (1) To establish a soil texture prediction model using GWR with a single bootstrap approach, (2) To test the model's reliability in predicting surface soil texture. The results of this study are in the form of a prediction model and a map of the spatial distribution of PSF on surface soil which can later be used as a basis for determining sustainable soil management and supporting precision agriculture.
Keywords:	GWR, Single Bootsrap, Soil particle-size fractions
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ENCRYPTION-BASED SECURE FRAMEWORK FOR SMART TRAFFIC MANAGEMENT USING FOG COMPUTING
Author:	SHWETA PANDEY, AVINASH KAUR
Abstract:	Inevitable traffic congestion has resulted from a growing population and their associated automobile requirements. In addition to diminishing the quality of everyday living, traffic congestion has long-term negative effects on an economy. Smart cities must be based on intelligent management systems, with traffic management at their centre. Smart traffic management (STM) depends on acquiring real-time data, processing, and organising the flow of traffic. This data can be acquired from different sources, including cameras, magnetic or piezo sensors, radar, and roadside units (RSU). Data collected from these various methods can be used to manage the traffic systems. However, the processing of data is central to smart traffic management. This analysis presents the application of fog computing to process the traffic data collected from different devices. The Fog computing architecture allows us to improve the latency, which enhances the overall performance of the system. Fog computing also reduces network usage compared to cloud computing. Here, a framework is presented, composed of various cameras and sensors that collect the data and transfer it to the fog nodes where it is processed and returned to the display. This would also assist in managing the traffic lights depending on the estimated congestion time.
Keywords:	Fog Computing, Smart Traffic Management, Fog Node, Edge Node, Cloud Node, Fog Computing Architecture
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	REAL-TIME RAY TRACING REFLECTIONS AND SHADOWS IMPLEMENTATION USING DIRECTX RAYTRACING
Author:	YOUNGSIK KIM
Abstract:	In traditional 3D games, techniques such as environment mapping and shadow mapping were used to simulate reflections and shadows due to the high computational load of ray tracing. However, recent advancements in technology have made real-time ray tracing possible, allowing for higher quality reflections and shadows compared to traditional methods. This paper proposes DirectX Raytracing (DXR) to achieve high-quality real-time reflections and shadows. To reduce the computational load of real-time ray tracing, we use the G-buffer from deferred rendering to compute only the information required for shadows and reflections, which is then combined to generate the final color. This paper verifies the effectiveness of our approach by comparing the performance of a DXR-based program with images produced using Unreal Engine 4's ray tracing capabilities. The results show that the proposed method provides high-quality graphics while minimizing computational load.
Keywords:	Reflection, Shadow, Real-time Raytracing, DirectX Raytracing, Performance Evaluation
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	DIGITAL MENU TRANSFORMATION: USABILITY TESTING APPROACH FOR THE FOOD AND BEVERAGE INDUSTRY'S
Author:	TANTY OKTAVIA1, STEPHEN KRISDY, MICHAEL NATHANIEL, JUSTIN JOE ADIWIDJAJA, JUAN MARCHELL KURNIAWAN, STEVEN ONG
Abstract:	Over the past few years, menu has shifted from the traditional physical menu into more advanced digitized form. Digital menu is a type of menu that utilizes the customer phone to show a restaurant's menu. The trend of using digital menu has attracted more and more restaurants to use digital menu instead of the traditional one. Since the transition is recent, a study is conducted to determine the influence of digital menus on customers. An online questionnaire was created in order to collect the data that is necessary to conduct the research. The data is collected from people within the region of Jakarta each with unique ages, and occupation. SmartPLS will be the software of choice to analyze the valuable result with a built-in method from SmartPLS, The Partial Least Squares - Structural Equation Modeling (PLS-SEM). From 129 respondents, this research took 100 samples and conducted an analysis on SmartPLS. The study showed that with the constant growing of technology on digital menu development, many people still prefer the usage of physical menu over digital menu with a variety of reasons. This leaves a big opportunity for restaurant owners to continuously improve their digital menu implementation in order to satisfy customers when using a digital menu.
Keywords:	Digital Menu, Physical Menu Smartpls, PLS-SEM
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	IMPROVEMENT OF STUDENTS ACHIEVEMENT BY USING INTELLIGENT TUTORING SYSTEMS - A BIBLIOMETRIC ANALYSIS AND REVIEWS
Author:	HAZEM A. ALRAKHAWI, NURULLIZAM JAMIAT, IRFAN NAUFAL UMAR, SAMY S. ABU-NASER
Abstract:	Intelligent tutoring systems (ITS) have emerged as a promising technology for improving students' achievement by providing personalized and adaptive learning experiences. This research presents a bibliometric analysis and review of the literature on the use of ITS for improving students' achievements. Using the creation and visualization of bibliometric networks software (VOSviewer), this investigation extended all studies from 2012 to 2022. In February 2023, a total of 88 publications were analyzed as recorded in the Scopus database, finding the most compelling topics addressed by the database using relevant keywords and criteria. The analysis aims to identify the research trends, influential authors, journals, institutions, countries, research areas and publications, and gaps in the literature related to ITS and student achievement. The results indicate a growing interest in the use of ITS for promoting student achievement, with a focus on personalized learning, artificial intelligence, and machine learning. The analysis identified influential authors, journals, institutions, countries, research areas and publications that have contributed to our understanding of the relationship between ITS and student achievement. However, the analysis also revealed several gaps in the literature, including the need for more research on the use of ITS for learners with diverse backgrounds or special needs and the implications of this research for the development of more effective instructional practices. The findings of this study can guide the development of more effective ITS and instructional practices and highlight areas for future research and development in this field. Overall, this study contributes to a better understanding of the potential of ITS in improving student achievement and promoting more efficient and effective learning.
Keywords:	Intelligent Tutoring Systems, Personalized Learning, Student Achievement, Adaptive Learning, Cognitive Science, Bibliometric Analysis
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ECONOMIC AND MATHEMATICAL MODELING OF INTEGRATION INFLUENCE OF INFORMATION AND COMMUNICATION TECHNOLOGIES ON THE DEVELOPMENT OF E-COMMERCE OF INDUSTRIAL ENTERPRISES
Author:	IGOR KRYVOVYAZYUK, IGOR BRITCHENKO, LIUBOV KOVALSKA, IRYNA OLEKSANDRENKO, LIUDMYLA PAVLIUK, OLENA ZAVADSKA
Abstract:	This research aims at establishing the impact of information and communication technologies (ICT) on e-commerce development of industrial enterprises by means of economic and mathematical modelling. The goal was achieved using the following methods: theoretical generalization, analysis and synthesis (to critically analyse the scientific approaches of scientists regarding the expediency of using mathematical models in the context of enterprises’ e-commerce development), target, comparison and grouping (to reveal innovative methodological approach to assessing ICT impact on e-commerce development of industrial enterprises), tabular, analytical and integral method (for summarizing the analysis results of enterprises readiness to implement ICT, ICT use in the activities of industrial enterprises of Ukraine and the analysis of e-commerce development), mathematical modelling (to build a regression model determining impact of changes in ICT use on the market share occupied by industrial enterprises), generalization (to determine promising directions of e-commerce developing of industrial enterprises). The implementation of a comprehensive approach to understanding the importance of ICT influence on e-commerce development of industrial enterprises will ensure acceleration of the digitalization of business processes, will contribute to the speed increase of enterprises response to customer requests, and increase the market share occupied by enterprises. A new vision of directions for developing e-commerce of industrial enterprises is suggested, which are determined by the need for enterprise rebranding, the development of e-commerce tools and technologies, the importance of outsourcing service automation and promotion of subscription trade. ICT is considered as integration factor that determines prospects for e-commerce development of industrial enterprises and contributes to increasing efficiency of online business management. Research results demonstrate that the use of economic and mathematical modelling is an important tool for assessing ICT impact, and its absence can negatively affect the accuracy and validity of online business management.
Keywords:	ICT, Online Business Management, E-Commerce, Economic And Mathematical Modelling, Decision-Making, Industrial Enterprises.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	PERFORMANCE ATTRIBUTES ANALYSIS OF SOFTWARE DEVELOPMENT COST MODEL WITH GAMMA FAMILY DISTRIBUTION CHARACTERISTICS
Author:	HYO JEONG BAE
Abstract:	In this study, the performance attributes of the NHPP-type software development cost model with Gamma family distribution characteristics widely known to be suitable for reliability studies were newly analyzed and evaluated. Also, after verifying the cost characteristics by comparing the proposed model with the Goel-Okumoto basic model, the optimal model was also presented. For efficient research, randomly collected software failure time data was used, and the estimation solution for the parameters of the proposed model was computed by maximum likelihood estimation. Conclusively, first, as a result of analyzing the m(t) function that affects the performance properties of the development cost model, the Rayleigh model and the Goel-Okumoto basic model were found to be efficient among the proposed models because the error value in predicting the true value was small. Second, when analyzing the development cost properties, the Rayleigh model was found to be an efficient model with excellent performance. Third, as a result of evaluating performance attributes, it was concluded that the Rayleigh model showed the best performance in this work. Therefore, if a software developer utilizes this research data, it can be used as a fundamental design guideline for the attributes analysis of development cost together with research on improving reliability quality.
Keywords:	Erlang, Log-Logistic, Rayleigh, Software Development Cost Model, Performance Attributes.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	A LITERATURE REVIEW OF SCENT TECHNOLOGY AND ANALYSIS ON DIGITAL SMELL TO CAPTURE, CLASSIFY, TRANSMIT AND REPRODUCE SMELL OVER INTERNET
Author:	SASEDHAREN CHINNATHAMBI, GOPINATH GANAPATHY
Abstract:	The Digital Smell Technology deals about Capturing, Classification, Transmission and Reproducing smell over the Internet. To improve on the experience we've had over the last few decades, our literature survey focuses on "e-Smell", that can transfer odours over the internet. The research is still in its early stages, smell appears to be an unrecognised medium and a new channel in multi-media, but the mystery of smell, combined with technological advancement, allows for the measurement and reproduction of odours.This study explores, how in the Digital Smell Technology, the smell can be transmitted over the Internet as static or streamed data. The technology reveals concepts in scientific disciplines such as chemistry, artificial intelligence, machine learning, data science, photonics, and not limited to electronics engineering. The goal of this study is to identify the limitations of Digital Smell Technology as a whole and the e-Nose in particular. The digital detection of various odours, digital transmission and reproduction of smell is now illuminating this technology. Our literature survey focuses on various dimensions about olfaction with respect to ongoing research and future challenges in Digital Smell. This paper presents the evolution of digital smell technology using e-nose methodology since 1950’s and discuss about the limitations of the existing e-nose mechanism and the lack of progression on capturing, transmission and reproduction of the smell.
Keywords:	Digital Smell Technology, Digital Scent Technology, e-Nose, Odour, Olfaction
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	HOW DARK WEB MONITORING CAN BE USED FOR OSINT AND INVESTIGATIONS
Author:	ALANOUD ALQUWAYZANI, RAWABI ALDOSSRI, M M HAFIZUR RAHMAN
Abstract:	The dark web is a hidden network of websites that cannot be accessed through regular search engines or browsers. It is often associated with illegal activities, such as the sale of illicit goods and services, human trafficking, and other criminal activities. Despite its illicit reputation, the dark web contains a wealth of information that can be utilized for open-source intelligence (OSINT) and investigations. This article explores how dark web monitoring can be utilized for OSINT and investigations. It discusses the ways in which dark web monitoring can be used to identify and track illegal activities on the dark web. Specifically, it examines the sale of illegal goods and services, the distribution of prohibited content, and the planning of criminal activities. By monitoring the dark web, law enforcement and security professionals can gain valuable insights into criminal activities and take appropriate action to prevent or mitigate them. However, dark web monitoring presents several challenges and limitations. The biggest challenge is the need for specialized knowledge and technical expertise to navigate the dark web safely and effectively. In addition, there is a risk of exposure to potentially harmful or illegal content, which can pose a risk to individuals or organizations who are not well-versed in the intricacies of the dark web. Overall, this article provides insights into the potential of dark web monitoring for OSINT and investigations. It emphasizes the need for caution and specialized knowledge to ensure that individuals and organizations can navigate the dark web safely and effectively. With the proper tools and expertise, dark web monitoring can be powerful for gathering intelligence and combating criminal activities.
Keywords:	Dark Web, Open-Source Intelligence, Investigation, Digital Crime, Cybercrime.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	HYBRIDML: FAULTY NODE DETECTION USING HYBRID LEARNING MODEL FOR DISTRIBUTED DATA CENTRES
Author:	ATUL V. DUSANE, DR. KRISHNAKANT. P. ADHIYA
Abstract:	The distributed systems are very effective when it deals with massive data processing. Nowadays, entire world generates high-dimensional data such as audio, video, image etc. To process such extensive data at a minimum is hard for a stand-alone machine, and this is a big challenge for the computer system to evaluate such data. The distributed framework is the solution for the process of such extensive data. Still, during the execution, some faulty or straggler nodes can increase the overall computation time to process data. However, to detect such straggler nodes, from large distributed systems are mandatory before assigning jobs to VM. Early identification of such faulty node can future save the overall computation time. In this paper, we proposed a hybrid machine learning model for detecting faulty nodes in large distributed machines using collaboration of reinforcement and supervised machine learning. The large Virtual Machine (VM’s) log data has been collected from the distributed environment and proceeded with reinforcement learning algorithm for module training and supervised machine learning for module testing. According to extracted features, reinforcement learning encompasses an activation function that generates the label for the respective node, whether healthy or faulty. In the testing phase, the natural world VM’s log data has been collected and evaluated with supervised machine learning classifiers. Several machine learning classification algorithms have evaluated and acquired the results. The SVM provides higher accuracy over the other machine learning classifiers with our reinforcement learning model.
Keywords:	Supervised Machine Learning, Classification, Faulty Node, Straggler Node Detection, Distributed System
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ANALYSIS AND CLASSIFICATION OF MALARIA INFECTED ERYTHROCYTES USING MICROSCOPIC IMAGES
Author:	SYED AZAR ALI, DR. S. PHANI KUMAR
Abstract:	Malaria is a member of a small group of diseases that can be highly severe. The presence of parasites in the environment might lead to sickness in the intestines. To visually identify the parasitemia, is a challenging task. The proposed solution categorizes and analyzes malaria impaired red blood cells. The parasite Plasmodium falciparum is the source of erythrocyte contamination. The system process the data in three phases/stages, the first phase preprocess the data by correcting the difference in luminance, the second phase performs segmentation of image pixels to detect erythrocytes. In the final phase a two stage classification model is used for identification of infected red blood cells. The system utilizes 3000 labeled images as the dataset for the purpose of training and 500 images for testing. The proposed system identifies infectious red blood cells with a Recall of 93% and 98.7% of Specificity. The model could demonstrate a recall of 77.9% and Specificity of 90.9% in identification of the infectious stage.
Keywords:	Malaria Parasites, Feature Extraction, Erythrocytes, AI Characterization, Automatic Analysis
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	THE INFLUENCE OF ELECTRONIC SERVICE QUALITY ON DIGITAL BANK APPLICATION
Author:	PALTI G. C. SINAGA, JAROT S. SUROSO
Abstract:	The advancement of technology and prevalence of smartphones in society influence the transition from offline to online purchases. The ability to carry out financial transactions from anywhere and at any time has become a necessity for the community, and banks have adapted to meet this demand. Customers' expectations for easy, safe, and more personalized services, as well as the ability to evaluate the quality of products and services, have led banks to go digital. This study utilizes two digital banks as research samples, with this digital bank being the one with the greatest number of downloads. However, compared to all other digital banks, these two have relatively lower ratings. This study employs a quantitative methodology and gathers data from 400 participants using an online questionnaire; all respondents had prior experience digital bank application. This study will investigate the factors that influence the loyalty of digital bank application users, as well as the amount to which each component influences the user's loyalty to digital bank applications. This study investigated the factors involved in utilizing the Electronic Service Quality model with variables consisting of ease of use, customer contact, reliability, responsiveness, security/privacy, and application design. This study employed Partial Least Squares Structural Equation Modeling (PLS-SEM), and the SmartPLS program was used to analyze the collected data. All variables except application design have positive and significant influence on customer loyalty. All factors have a positive and significant influence on customer satisfaction. Additionally, the satisfaction determined to have a positive and significant influence on customer loyalty. This research is presumed to assist Digital Bank in enhancing and improving their banking applications and improving their services, so that customer satisfaction and loyalty keep on improving.
Keywords:	Banking Industry, Customer Loyalty, Customer Satisfaction, Digital Bank, Eelectronic Service Quality.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	POSITIVE INTERVENTION TECHNIQUE FOR A COMPUTER VIRUS
Author:	ABDULRAHMAN ALKANDARI, MOHAMMAD ALAHMAD, NAYEF ALAWADHI, ABDULLAH ALSHEHAB, ADEL ALFOUDERY
Abstract:	Negative connotations have generally been associated with computer viruses. In particular, popular belief holds that computer viruses are harmful and many people tend to use them for malicious reasons. However, they are not entirely bad as its known. They are codes that can either be used profitable or harmfully. This bias in existing literature has necessitated this research that aims to evaluate the possibility of using viruses for beneficial purposes while simultaneously improving storage and computing efficiency. First, related work on the subject is reviewed. Second, approaches to accessing virus code for editing purposes are reviewed. Third, beneficial aspects of viruses pioneered by Fred Cohen are reviewed. Fourth, computing and storage efficiency with use of beneficial viruses are reviewed. Lastly, Anti-Virus techniques that have been adopted in dealing with contemporary computer viruses have been reviewed. In this research, we present the possibility of changing computer virus from harmful to useful.
Keywords:	Computer Virus, Beneficial Virus, Source Code, Self-Replication, Metadata.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	METAHEURISTICS WITH DEEP CONVOLUTIONAL NEURAL NETWORK FOR CLASS IMBALANCE HANDLING WITH ANOMALY DETECTION IN INDUSTRIAL IOT ENVIRONMENT
Author:	NENAVATH CHANDER, MUMMADI UPENDRA KUMAR
Abstract:	The advancements of industrial Internet of Things (IIoT) have brought substantial value and accessibility to the industry. At the same time, it is followed by various security risks involving anomalies in the gathered data. Anomalies could emerge in the system because of several reasons namely software and hardware malfunctions, or a cyber-attack. The major problems in designing an effectual anomaly detection system include complexity in different anomaly definitions in various domains, defining normal region, normal behavior variation over time, the noise presence in the datasets, and lack of suitable datasets. Furthermore, Class imbalance is the term utilized for data having minority and majority classes. The spectrum of class imbalance ranges from “slightly imbalanced” to “rarity”. In a majority–minority classification problem, class imbalance in the data can drastically skew the classifier performance, presenting a prediction bias for the majority class. This study develops an optimal Deep Convolutional Neural Network for Class Imbalance Handling Anomaly Detection (ODCNN-CIHAD) model. The proposed ODCNN-CIHAD technique majorly focuses on two major processes namely class imbalance data handling and anomaly detection. At the initial stage, the ODCNN-CIHAD technique follows min-max data normalization technique to convert the input data into compatible format. In addition, the ODCNN-CIHAD technique designs a group teaching optimization algorithm (GTOA) with SMOTE technique for handling class imbalance data. Also, the DCNN approach was applied for the recognition and classification of anomalies that exist in the IIoT data. Finally, the gorilla troops optimizer (GTRO) approach was exploited for optimum hyperparameter tuning of the DCNN approach. The experimental validation of the ODCNN-CIHAD technique is carried out utilizing benchmark dataset and the outcomes are inspected under various measures. The comparison study highlighted the improved performance of the ODCNN-CIHAD system on existing approaches.
Keywords:	Security; Anomaly detection; Industrial Internet of Things; Deep learning; SMOTE technique; Class imbalance data
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	DESIGNING A UNIVERSAL DIGITAL TWIN OF AN OBJECT BASED ON A HYBRID NEURO-FUZZY COMPUTER
Author:	SYRYAMKIN V.I., GORBACHEV S.V. , KLESTOV S.A. , ABRAMOVA T.V., UVAROV N.A. , GUTSAL V.A., MEHTIEV A.D., SERIKOV T.G., MANANKOVA O.A, UTEGENOVA A.S.
Abstract:	Designing a universal digital twin of an object based on a hybrid neuro-fuzzy computer is a topic that has been widely studied in recent years as it provides many benefits such as improving performance, reducing downtime, and increasing efficiency. This paper presents a method for creating a hybrid computational neuro-fuzzy double of an object. The goal is to create a digital twin of the object that can be used for monitoring, simulation, and prediction of the object's performance, while also utilizing the strengths of both neuro-fuzzy computation to improve the accuracy and robustness of the digital twin. The proposed method shows potential for improving the performance and efficiency of physical objects. In the paper, the proposed method forms an electronic model - a digital double of the synthesized object. The proposed calculator is superior to a similar one in several advantages.
Keywords:	Digital Twin, A Hybrid Neuro-Fuzzy Calculator, Computer-Aided Design System, Neural Network Algorithms, Simulation, Neuro-Fuzzy Computer.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ABUNASER - A NOVEL DATA AUGMENTATION ALGORITHM FOR DATASETS WITH NUMERICAL FEATURES
Author:	BASEM S. ABUNASSER, SALWANI MOHD DAUD, IHAB S. ZAQOUT, SAMY S. ABU-NASER
Abstract:	This research paper introduces Abunaser, a novel data augmentation algorithm for numerical datasets. Abunaser is designed to address the challenge of overfitting in machine learning models when working with small numerical datasets. We evaluate the effectiveness of Abunaser in improving the performance of machine learning models on numerical datasets and compare it with other commonly used data augmentation techniques. Our results show that Abunaser can effectively increase the size of the dataset and improve the performance of machine learning models across different types of tasks, including classification, regression, and clustering. We also investigate the sensitivity of Abunaser to different parameters, such as the size of the dataset and the number of features. Additionally, we provide insights into the underlying mechanisms of Abunaser and how it affects the distribution and structure of the augmented data. However, we acknowledge some limitations of our research, including the dataset characteristics and computational requirements of Abunaser. Overall, our study suggests that Abunaser is a promising data augmentation algorithm for numerical datasets and has the potential to improve the performance of machine learning models in various applications.
Keywords:	Dataset, Augmentation, Machine learning, supervised models, Deep Learning
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	INTELLIGENT TOUCHLESS SYSTEM BASED ON GESTURE RECOGNITION
Author:	AISWARYA BABU, ZAHIRIDDIN RUSTAMOV, SHERZOD TURAEV
Abstract:	In our rapidly advancing technological era, every industry is experiencing its revolution. As we navigate the challenges the current pandemic presents, there is a heightened interest in solutions that facilitate social distancing and contactless interactions. To address this challenge, we propose the development of an interactive and innovative platform that allows users to navigate through hand gestures. This touchless system can be customized to meet various needs and utilizes a set of standard hand gestures for simplicity and ease of use and can be implemented in multiple sectors such as airports, banking, retail, restaurants, and so on. To demonstrate the system's potential, we have created a mobile food ordering application that uses hand gestures as the primary means of interaction and uses a set of standard hand gestures to promote simplicity, familiarity, and user accessibility. This study will develop a mobile food ordering system to illustrate the proposed gesture-based touchless system. To build our gesture recognition model, we collected a dataset of common hand gestures by scraping images from the web. We then trained our models using the Efficient Net-Lite [0-4] algorithms, leveraging transfer learning and pre-trained deep learning models to reduce computational demands. We utilized transfer learning and pre-trained deep learning models to reduce the time and computational resources required for training. The trained models were evaluated using the mean average precision (mAP) and inference time and then converted into a lightweight format, TensorFlow Lite, for use on mobile devices such as kiosks for the mentioned scenario. Our evaluation results revealed that all the trained models achieved an mAP of 82% or higher, with the most complex model, EfficientNet-Lite4, reaching 87%. However, the inference time for the trained models was significantly longer, ranging from one to ten seconds. To balance performance and inference time, we chose the EfficientNet-Lite0 model with an inference time of just half a second for our hand gesture-based touchless system. This model provides an adequate level of accuracy for our hand gesture-based touchless system while minimizing any lag or delay that could impact user experience. In summary, our proposed system is a cutting-edge, user-friendly solution that meets the need for contactless interactions and social distancing. Using standardized hand gestures, we have created a platform that is intuitive and accessible for users. Our system has the potential to offer significant benefits across a wide range of industries and applications in the modern era.
Keywords:	Machine Learning, Artificial Intelligence, Hand Gestures, Contactless, Gesture-Based
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	DIGITAL LITERACY ON CURRENT ISSUES IN SOCIAL MEDIA: SOCIAL MEDIA AS A SOURCE OF INFORMATION
Author:	BURHANUDDIN ARAFAH, MUHAMMAD HASYIM
Abstract:	The digital age has shifted human information access to technological-based media. A smartphone, for example, is a technological product that provides quick access to the current information or news available virtually so that all Smartphone users can immediately get the information or news. The advancement of this technology-based communication allows internet users to receive promptly, send a comment, and share news with other users through social media such as WhatsApp, Facebook, and Instagram. Since social media, as part of internet media, provides the latest hot news, this study highlighted the use of social media in disseminating recent issues. The data were taken from the written news or information on WhatsApp, Facebook, or Instagram, and the questionnaires were distributed to internet users via social media. More particular research problems are what recent issues are disseminated on social media and how the digital literacy of internet users is related to the current issues posted on social media. Based on the result of the data analysis, it is found that 90.03 per cent of internet users access information through social media, and 81 per cent spend time searching for information on social media. Concerning the percentage of users of each social media, the findings denote that 38.4 per cent of users access Facebook, 20.2 per cent access WhatsApp, 18.4 per cent access YouTube, 8.3 per cent access Twitter, and 8.3 per cent access Tiktok. News of covid-19 vaccination and religious intolerance are the current issues that social media users access. However, the digital literacy of internet users is the main issue that hoax news about Covid-19 vaccination and religious intolerance are disseminated through social media.
Keywords:	Social Media, The Gate Of Information, Recent Issues, Digital Literacy
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	COOPERATIVE TRUST FRAMEWORK BASED ON HY-IDS, FIREWALLS, AND MOBILE AGENTS TO ENHANCE SECURITY IN A CLOUD ENVIRONMENT
Author:	HICHAM TOUMI, FATIMA ZAHRA FAGROUD, KHADIJA ACHTAICH, FATIMA LAKRAMI, MOHAMED TALEA
Abstract:	Cloud computing has indeed become a popular method for hosting and delivering internet-based services due to its efficiency and scalability. However, as with any technology, there are inherent security risks associated with it. Organizations must carefully evaluate the security measures provided by their chosen cloud provider and implement additional security measures as needed to protect their data and applications from potential attacks. Security is a crucial concern in cloud computing, and ensuring client satisfaction requires transparency, reliability, and increased security measures. Preventing and mitigating the impact of potential intrusions is a top priority, given the dynamic nature of cloud computing environments. In addition, effective resource protection and recovery must be in place to ensure business continuity without relying on external intervention. addressing the self-healing of cloud security requires the utilization of fundamental aspects of autonomic computing in the cloud. The strong alignment between autonomic computing systems and multi-agent systems allows for the creation of an intelligent cloud architecture that can effectively support autonomic aspects. Therefore, a cooperative Hybrid Intrusion Detection System (Hy-IDS), Mobile Agents, and Firewalls framework have been proposed to counter security attacks in this environment. Our solution offers an extra layer of preventative and protective security measures that not only detects known intrusions but also detects variations of multiple known attacks and previously unknown attacks.
Keywords:	Cloud Computing, Virtualization, Security, Firewalls, IDS
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	EMPIRICAL INVESTIGATIONS TO A NEIGHBORING IDEAL MULTICAST SCHEME FOR MOBILE AD-HOC NETWORKS
Author:	DR. THOTAKURA HARITHA, DR. SURYA PRASADA RAO BORRA, DR. A. GEETHA DEVI, PRAVEEN TUMULURU, DR.K. VIDYA SAGAR
Abstract:	Ad-hoc mobile networks are made up of mobile nodes that are randomly and dynamically placed so that their connections to one another can change at any time. Such an ad-hoc network routing protocol's main objective is to construct an accurate and effective route between two nodes so that messages may be delivered promptly. Instead of sending several copies of a packet over the same area of the network or sending packets to clients who don't want them, multicasting is the practice of sending a single copy of a packet to all of the clients that want it. By using user-multicast trees and dynamic logical cores, the Adhoc Multicast Routing Protocol (AM Route) offers a revolutionary method for resilient IP Multicast in mobile ad-hoc networks. Only group senders and receivers are used as tree nodes in order to establish a bi-directional, shared tree for data dissemination. On the User-multicast tree, neighbours are connected by means of unicast tunnels. As a result, network nodes that are not interested in or capable of multicasting do not need to implement AM Route, and only group senders and receivers are responsible for group State Cost. Additionally, even in the event of a changing network topology, using tunnels as tree connections indicates that tree structure does not need to change, which lowers signaling traffic and packet loss. As a result, the underlying Unicast protocol serves this purpose and AM Route is not required to monitor network fluctuations. Since AM Route doesn't need a particular Unicast routing protocol, it may work without a hitch over distinct domains using various Unicast protocols. The transient loops in the mesh production have been addressed. Additionally, in order to increase the protocol's effectiveness, we introduced the dynamic core migration approach by employing a timer that periodically switches the current core node.
Keywords:	Adhoc, AM Route, Ideal, Scheme, route.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ADVANCING UNIVERSITY LEARNING WITH EMOTIONAL INTELLIGENCE AND MODEL-DRIVEN ENGINEERING: DEVELOPMENT AND EVALUATION OF A TEST PLATFORM
Author:	OUZAYR RABHI, MOHAMMED ERRAMDANI, Saida FILALI
Abstract:	This paper presents a novel approach that combines emotional intelligence and model-driven engineering to develop a personalized test platform for university students, filling a gap in the literature. Our approach integrates emotional intelligence assessment with adaptive testing algorithms to provide customized feedback for enhancing students' emotional and academic outcomes. While emotional intelligence and model-driven engineering have been extensively researched separately, there is a lack of research that combines these two fields to provide personalized, data-driven, and emotionally intelligent assessment and feedback systems for students. To establish the novelty of our approach, we evaluated the system's effectiveness in a pilot study involving university students, which resulted in the creation of new knowledge. Our study demonstrated the effectiveness of integrating emotional intelligence assessment and MDA to enhance students' emotional and academic outcomes, contributing to the advancement of the integration of these two fields. This novel approach has the potential to revolutionize the field of education by providing personalized, data-driven, and emotionally intelligent assessment and feedback systems.
Keywords:	Emotional intelligence, MDA, Psychological test, PIM to PSM,
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	EFFICIENT IOT-BASED CLOUD COMPUTING FRAMEWORK FOR SECURE DATA STORAGE USING MACHINE LEARNING ALGORITHM
Author:	RUPALI S. PATIL, AMINA KOTWAL, SWATI S. PATIL
Abstract:	Cloud computing is a widely used technology that has changed the way people and organizations store and access information. This technology is versatile, and extensive amounts of data can be stored in the cloud. However, with the development of cloud computing, it is also faced with many difficulties, cloud computing security has become the leading cause of impeding its development. Cloud computing security has become a hot topic in industry and academic research. As a consequence, the security of data stored in the cloud serves as a key concern for cloud consumers due to ongoing hacking incidents in the cloud. This work used encryption with access management because authenticities, anonymity, and security over accessibility are mandatory. Accordingly, the article proposed a machine learning-based method for secure data storage in the cloud. Initially, the data is compressed using the Huffman algorithm, which minimizes text data size and storage, resource use, or transmission power. Accordingly, the compressed data are encrypted using a novel cryptographic technique. This method encrypts the data before uploading it onto the cloud. Subsequently, the malicious intention in the cloud platform is identified by proposing a Weighted Chimp Algorithm optimized Gaussian Kernel Radial Basis Function Neural Network. This malicious code can be spread through infrastructures in the cloud platforms and pose a great threat to users and enterprises. The proposed method accurately detects malicious code in the cloud. The proposed work is implemented using Python software. The proposed method is compared with the other existing methods like Fully Homomorphic Encryption (FHE), Ciphertext Policy-Attribute based Encryption (CP-ABE), and Quasi Modified Levy Flight Distribution Reversed Sheamir Algorithm (QMLFD-RSA). Accordingly, the proposed method outperforms these existing methods. The result revealed that the deduplication rate, throughput, cipher text and encryption time of the proposed method produce higher performance than the existing methods, ie) the deduplication rate for the proposed method is 94% and the outcome of the work proved that the proposed work produces better security than the other existing research respectively. This hybrid technique provides the user to get an advantage from retrieved information in a protected manner.
Keywords:	Cloud Computing, Security, Data Storage, Huffman Algorithm, Data Compression, Malicious Behaviour, Gaussian Kernel Radial Basis Function Neural Network, and Weighted Chimp Algorithm.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	PATHOLOGY CARDIAC MONITORING STUDY OF THE PHONOCARDIOGRAM SIGNAL
Author:	DEBBAL IMANE, HAMZA CHERIF LOTFI, BAAKEK YETTOU NOUR EL HOUDA
Abstract:	In this paper, we will discuss the efficiency of the Fast Fourier transform (FFT) and the Short-Time Fourier Transform (STFT) to distinguish cardiac pathological signals, along with following the severity evolution of diverse diseases through three selected features. The cardiac signals analysed and previously classified via some clinical data will be arranged into three main classes or groups: a group of signals containing neither clicks nor murmurs and having a similar morphology, a second group of signals containing only clicks (reduced murmurs), and a third group of signals with a significant murmur. The features that we are going to define from each technic will help us in this sense to classify the different signals analysed in one of the mentioned groups. We will then extract the same features from a fourth group of phonocardiogram (PCG) signals suffering from murmur with different severity levels. In the end, we will discuss the accuracy of these features with the Energetic Ratio (ER) parameter and a K-Nearest Neighbor classifier in terms of classifying phonocardiogram (PCG) signals according to their pathological origin and cardiac severity level. An accuracy of 99.2% is achieved when using a combination of time and spectral features (frequency band (FB), frequency extent (∆F), time extent (∆T)) to classify the PCG signals in the three main groups and a 98.9% accuracy when ranking signals according to their severity level (Light, Moderate, Severe). The main aim of this paper is to proceed with the use of the FFT and the STFT technics to obtain information likely not only to discriminate the three groups' cases but also to detect the level (or degree) of severity in the same studied pathology as well. Thus, the intent of this study on phonocardiogram (PCG) signals is to follow the evolution of the pathology at different levels and identify each severity degree via the extracted features, which makes the originality of this paper. These results can only help the clinician to make his decision with serenity.
Keywords:	Phonocardiogram, Normal, Pathological, Classification, Discrimination, Severity, FFT, STFT, Spectral, Time Extency, Frequency Extency.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	FINETUNED ROBERTA ARCHITECTURE FOR MOOCS EVALUATION USING ADVERSARIAL TRAINING
Author:	MUNIGADAPA PURNACHARY, T ADILAKSHMI
Abstract:	Massive Open Online Course is an online learning platform that can provide the opportunity to the learners to gain knowledge in a particular domain or subject. During recent days, MOOCs grabbed the attention of learners and provide quality education for free. As many MOOC providers are offering same courses, Choosing the best one out of all available MOOCs become very challenging task. MOOCs Evaluation plays an important role in finding the sentiment of the learners through the reviews and helps the MOOC Providers to improve the curricular quality. This paper proposed MOOC-RoBERTa, a sentiment analysis architecture that can evaluate MOOCs using Student reviews. Firstly, we prepared a balanced MOOC Reviews Dataset containing 13200 reviews. Secondly finetuned a RoBERTa Model for sentiment analysis on MOOCs. Next, trained the proposed model by applying Adversarial attacks approach to gain domain specific knowledge and test the model to find the sentiment of the learners. Finally, we compared the performance of the proposed model with different variants of other transfer learning models like BERT, Albert, XLNet. The experimental results demonstrate that the proposed model outshines the state-of-the-art methods by achieving the accuracy of 96.6% on MOOC Reviews Dataset.
Keywords:	MOOCs, Sentiment, Transfer learning, BERT, XLNet, Adversarial Attack, Accuracy
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	HYBRID FEATURE SELECTION MODEL BASED ON RFE AND MRMR ON ANXIETY DISORDER DATASET
Author:	PRAJESHA T. M., S. VENI
Abstract:	According to the reports from World Health Organization large volume of people are suffering from some kind of anxiety disorders. Among which Generalized Anxiety Disorder (GAD) is the most common one. It is very necessary to identify the anxiety disorders in beginning stage otherwise it leads to medication. Aim of this paper is to create a model to find the generalized anxiety disorder with high accuracy with lesser number of features. It describes a method to find the most important features to determine GAD. A hybrid feature selection approach is proposed which combines wrapper and filter feature selection methods to find out the best features to predict anxiety disorder. A hybrid model removes the biases that may exist while using single models. This hybrid feature selection approach is tested with four classification algorithms such as Random Tree, REP Tree, JRip, LogitBoost. As compared to the previous works this model makes the prediction more accurately with seven attributes. The dataset used for the analysis is collected from an online survey among persons above eighteen years. Analysis shows that the performance has improved after feature selection and among the classification algorithms LogitBoost gave better performance measures. It also found that the questionnaires in Hospital and Anxiety Depression Scale is well suited than other measures for finding the generalized anxiety disorder.
Keywords:	Generalized Anxiety Disorder, Recursive Feature Elimination, Feature selection, Minimum Redundancy and Maximum Relevance, Classification
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	INTERNET OF THINGS FOR EFFORT ESTIMATION AND CONTROLLING THE STATE OF AN ELECTRIC VEHICLE IN A CYBER ATTACK ENVIRONMENT
Author:	BADDU NAIK BHUKYA, VUTUKURI SARVANI DUTI REKHA, VENKATA KRISHNAKANTH PARUCHURI, ASHOK KUMAR KAVURU, KADIYALA SUDHAKAR
Abstract:	The Internet of Things (IoT) lets millions of smart devices sense, gather, process, and exchange data to provide intelligent services. IoT-based communication infrastructure allows cyber-physical devices like electric cars to sense, monitor, and be controlled remotely. IoT cannot explore these uses due to cyberattacks on traditional communication infrastructure. This paper suggests an algorithm for monitoring and managing electric vehicles via the Internet of Things while preventing false data-injection attacks. First, a vision-equipped fully autonomous electric vehicle state-space model is described. Smart sensors and actuators in the Internet of Things infrastructure watch and adjust system states to compensate for the long distance between the electric vehicle and the control centre. Vehicle sensing data is sent to a central command centre via a vulnerable communication route. The mean square error principle yields the best state estimation method for visualising vehicle states. An optimal control algorithm manages car states using semi-definite programming. Simulations demonstrate how well the proposed algorithms can foresee and control vehicle states.
Keywords:	Internet Of Things (Iot), Cyber-Attacks, Electric Vehicles, Communication Network, Control Center.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ONTOLOGY MATCHING USING DEEP LEARNING
Author:	ZAKARIA HAMANE, AMINA SAMIH, ABDELHADI FENNAN
Abstract:	Ontology matching is a critical task in knowledge representation and integration, with numerous applications in various domains. Deep learning methods have shown promising results in improving the accuracy and efficiency of ontology matching. However, there is a lack of comprehensive analysis and classification of these methods in the literature. In this paper, we conducted a systematic literature review of ontology matching using deep learning methods, covering articles published between 2005 and 2022. Our analysis includes a trend analysis of the articles, a framework for classifying them, and a detailed classification of the articles based on the deep learning method used.
Keywords:	Ontology Matching, Ontology Alignment, Literature Review, Deep Learning, Word Embedding
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	OPTIMAL PLACEMENT AND ADAPTIVE CONTROLLING FOR DOUBLY-FED INDUCTION GENERATOR INTEGRATION IN SMALL SIGNAL STABILITY
Author:	ANJALI V. DESHPANDE, DR. V. A. KULKARNI
Abstract:	The optimal location of DFIG improves the small signal stability of wind integrated power system. Also when parameters of PSS are optimized, it helps to improve the small signal stability. In this paper along with optimal placement of DFIG, PSS parameters are optimized which further improves the small signal stability of the power system. The enhancement in small signal stability with optimal location and optimized parameters is shown in this paper using Eigen value analysis. For locating the wind farm in a multi machine system at its optimal position, an Eigen value index (EI) is used. The Eigen value objective function is used for obtaining optimal parameters of PSS. Analysis of the integrated optimization of DFIG location and PSS parameters shows a more effective small signal stabilization in the power system. The placement of DFIG unit at different bus locations reflects in variation of damping factor and movement of the Eigen values. The simulation of a two-area network with four machines, shows that the placement of DFIG unit at bus 4 with optimal PSS parameters resulted into a shift of -0.4 in the real component compared to the previous stabilizing methods. The index (Iss) in terms of damping factor which is a measure of stability, is observed to be improved by 5%when the load is varied by 10%, as compared to the previous individual PSI and WOA methods. Thus the dual optimization, with the use of PSIs for optimal location of DFIG and WOA algorithm for optimal parameters of PSS, used here has proved to be a novel method of improving small signal stability.
Keywords:	Small Signal Stability, DFIG Location, PSS Parameters, Eigen Value Analysis, Probability Sensitivity Index.
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	SUPPLY CHAIN RISK ASSESSMENT WITH FUZZY LOGIC APPLIED TO THE FAILURE MODE AND EFFECT ANALYSIS METHOD
Author:	LAKHOUIL HATIM, SOULHI AZIZ
Abstract:	The context of crisis caused by unpredictable events and lack of visibility has given rise to a new concept: the uncertainty , induced by the supply chain disruptions , As a result, decision-making has become increasingly a complicated operation, which has motivated supply chain managers to look for others new tools adapted to unclear circumstances to measure risks and define priorities . in this work, supply Chain risk assessment has been modelled using the famous FMEA method, which has become a method widely used by industrialists, however the originality of this work lies in its merge with the fuzzy approach initiated by LOTFI ZADEH in 1965, and which will allow us to pass from real data to linguistic sentences that are better understood by humans brains and also close to reality by introducing the membership functions and inference rules that will allow us to measure risk through defuzzification, which will support decision-makers in judging all the alternatives and scenarios that can be found in their analysis.
Keywords:	Supply Chain Uncertainty , Decision-Making , FMEA , Fuzzy Logic , Risk Assessment
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	PREDICTING THE SEVERITY OF NEW SARS-COV-2 VARIANTS IN VACCINATED PATIENTS USING MACHINE LEARNING
Author:	MEROUANE ERTEL,AZEDDINE SADQUI, SAID AMALI, INTISSAR MAHMOUDI, YOUNES BOUFERMA,NOUR-EDDINE EL FADDOULI
Abstract:	Given the increasing number of COVID-19 cases and the risk of new variants, early prediction of disease severity in critical care patients is essential to optimize treatment options. In this study, we set up an experiment on 236 patients infected with COVID-19 and hospitalized at the Sidi Said hospital in Meknes, Morocco. This work proposes a new multivariate classification model to predict which patients admitted to hospital with COVID-19 will require special care (oxygen therapy, intensive care, resuscitation) or will die following an abrupt deterioration in their state of health. This model will help healthcare professionals (doctors) make decisions about recommending appropriate medical treatments to patients. A comparative study of different multivariate machine learning algorithms (Support Vector Machine (SVM), K-nearest neighbor (KNN), Decision Tree (DT) and Random Forest (RF)) is also presented in this article. The result obtained shows that the SVM classifier is a reliable, powerful and efficient algorithm to predict the level of risk of patients contaminated with COVID-19.
Keywords:	Covid-19; Clinical Decision Support; Machine Learning; Ordinal Classification, Multi-Class Classification; Personalized Medicine
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	COMPREHENSIVE ANALYSIS OF MACHINE LEARNING ALGORITHMS TO DETECT ALZHEIMER’S DISEASE USING PREDICTOR FACTORS
Author:	DEEPTHI K. OOMMEN, J. ARUNNEHRU
Abstract:	Alzheimer's disease (AD) is an irrecoverable encephalopathy. The condition diminishes intellectual capacity and causes memory loss. To collect pertinent images and train to identify AD and its phases, computer-aided diagnostic approaches with image retrieval have created a new perspective in MR imaging. Although computer-assisted techniques have achieved considerable research advancements, the viable diagnosis method available in clinical practice still needs improvement. Recently, there has been a proliferation of high-scale results displayed by more advanced machine-learning approaches in various domains. This paper focused on classifying the subjects who potentially can have Alzheimer's disease with machine learning and Deep learning techniques. The models encompassed OASIS dataset for diagnosing the disease. Clinicians can diagnose and classify these disorders using the proposed classification approach. The computational algorithms can help practitioners in reducing the average annual fatality rates of Alzheimer's disease by early diagnosis. Extensive research was done to find the significant predictor parameters and measure how well the model works using performance metrics, which gives its uniqueness. The performance of the extra tree classifier was superior when compared with other ML models, with an accuracy of 86%. The Deep Neural Network (DNN) acquired an optimum accuracy in the binary classification with 92%.
Keywords:	Alzheimer’s Disease, Machine Learning, Brain Disorder, Predictor Factors, OASIS
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	ENERGY EFFICIENT DYNAMIC NEIGHBOR DISCOVERY SCHEDULE ALGORITHM FOR WIRELESS SENSOR NETWORKS
Author:	DR SAGAR MEKALA, DR KALAIVANI D, RAMA RAO TANDU, KALANGI PRAVEEN KUMAR, DR G.RAJAVIKRAM, VADAPALLY PRAVEEN KUMAR
Abstract:	The recent rapid technological developments of Internet of Things (IoT) brought in Wireless Sensor Networks (WSNs) are more and more widely used in many applications. A WSN usually consist of a deployment of thousands of tiny nodes and a base station (BS) on an interested phenomenon. To such structured network, each node is able to gather physical information and transmit to a sink node. To communicate data from node to BS, coordination between nodes and energy efficient communication is needed. Neighbor discovery process performs vital responsibility in WSNs, due to resource constrained tiny devices in network and to maximizing the network lifetime. In recent years, many neighbor discovery schemes developed to minimize energy consumption as much as possible and at the same time make sure that discovery latency as small as possible. A node typically has two options for learning about its neighbours: synchronously and asynchronously. Given the resources available and the dynamic nature of networks, synchronization between nodes is a challenging issue. Many asynchronous neighbor discovery protocols are proposed to address issues and challenges of discovery of neighbors in WSNs with the help of probabilistic, deterministic, and quorum based approaches. In this paper, we adapt the concept of dynamic schedule based on block design and combinatorial methods for asymmetric neighbor discovery. First, we summarize need of neighbor discovery and source of power utilization in the neighbor discovery process, subsequently discussed about the difficulty of designing block design. To improve the energy utilization, we construct dynamic schedule mechanism for asymmetric neighbor discovery. We analyze the worst-case discovery latency in our proposed model with representative algorithms in the parallel research. Experimental results of our simulation represent that the worst case discovery delay significantly better than that of traditional algorithms.
Keywords:	Energy Efficient, Neighbor Discovery, WSN, Block Design, Discovery Latency
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text

Title:	IMAGE SEGMENTATION BASED ON TINT USING DATA MINING TECHNIQUES
Author:	P ANIL KUMAR, DR.ASHA LATHA BANDI, B V N V KRISHNA SURESH, DR. A. JAGADEESWARA RAO, DR. CH SURESH BABU, DR. SHAHEDA NILOUFER, KOTESWARA RAO KODEPOGU
Abstract:	Segment-based image analysis techniques are becoming increasingly crucial for creating and updating geographical information, mostly due to advancements in satellite imagery's spatial resolution. This paper provides a unique unsupervised K-means clustering approach for segmenting images based on colour data. We didn't utilize any training data in this. Two stages make up the overall project. Before the areas are classified into a set of five classes using the k-means clustering technique, the colour separation of the satellite picture is first improved using de-correlation stretching. By skipping feature calculation for each pixel in the picture, the computational cost of this two-step procedure can be decreased. Despite not being often utilised for picture segmentation, the colour contributes gives a high discriminative power of regions present in the image.
Keywords:	Resolution, Image Segmentation, Data Mining, Cluster, Color
Source:	Journal of Theoretical and Applied Information Technology 31^st May 2023 -- Vol. 101. No. 10-- 2023
Full Text