		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

February 2023 | Vol. 101 No.4

Title:	MALIMG2022: DATA AUGMENTATION AND TRANSFER LEARNING TO SOLVE IMBALANCED TRAINING DATA FOR MALWARE CLASSIFICATION
Author:	IKRAM BEN ABDEL OUAHAB, LOTFI ELAACHAK, YASSER A. ALLUHAIDAN3, MOHAMMED BOUHORMA
Abstract:	Data augmentation is creating new images by transforming old ones. Its used to solve imbalanced image classification problems in many domains. Usually, data augmentation is used when we are unable to get more data for underrepresented classes. So, data augmentation techniques help us to increase the size of training data in order to avoid any bias in the classifier. This papers main contribution is to develop a balancing tool for any imbalanced multiclass database. Then, we use this approach in application to the Malimg database to improve its effectiveness and speed to solve imbalanced data problems. As a result, we generated 2 versions of the Malimg database namely Malimg2022 (Large and XXLarge), also we make the (Balance Me) application that can balance any database using augmentation techniques. These new versions are balanced, having the same number of samples per class using data augmentation with different transformation techniques. From a technical point of view, Zero-day malwares are none than old ones with few modifications, so data augmentation could be seen as a simulation of new malware variants that should be detected effectively. Finally, the new balanced data were evaluated using transfer learning models. So, the generated databases didnt show an overall improvement in the whole classification task, however, we have found some improvements related to some malware classes. Here, we can say that the use of data augmentation to balance data isnt always a good choice.
Keywords:	Cybersecurity, Image augmentation, Malware classification, Transfer Learning, Deep Learning.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	DIGITAL METHODS FOR UNERRORIC OF THE ONBOARD LOCATOR OPERATOR WORK QUALITY CONTROL
Author:	ADELIYA BUROVA, ALEXANDER SHEMYAKOV, ANDREY SOROKIN
Abstract:	The questions connected with research of additional possibility of estimation automation of the onboard locator operator means work results directly onboard are considered. The aim of the study is to develop the concept and methodology of hardware-software support of multilevel quality control of the operator of onboard locating means in the training mode of their functioning onboard the ship. During the study, methods of hardware-software modeling of deductive algorithms for digital signal processing were used. The proposed and developed principles and digital methods of multilevel quality control of such work in the training mode of operation of onboard locating means are described. The structural scheme of multilevel evaluation of the results of training of the operator of these tools during their functioning in the training mode is given.
Keywords:	Hardware-Software, Onboard Locating Means, Multilevel Quality Control, Training Results Evaluation, Training Mode, Digital Signal Processing.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	USE OF CHATBOTS IN E-COMMERCE: A COMPREHENSIVE SYSTEMATIC REVIEW
Author:	JAVIER GAMBOA CRUZADO, CHRISTOPHER MENENDEZ-MORALES, CARLOS FRANCO DEL CARPIO, JEFFERSON LOPEZ GOYCOCHEA, ALBERTO ALVA AREVALO, CALEB RIOS VARGAS
Abstract:	The advance of Chatbots nowadays presents a great number of techniques for its subsequent implementation in e-commerce, thus diverting from the traditional user experience that is normally had when making purchases when browsing the Internet. The objective of the research was to know the state of the art about Chatbots and their impact on E-Commerce. A systematic literature review (SLR) was conducted based on B. Kitchenham and S. Charters [76] from 2017 to 2021. The search strategy identified 233 085 papers from digital libraries such as Scopus, IEEE Xplore, Google Scholar, Springer, ACM Digital Library, ProQuest, Taylor & Francis Online and Wiley Online Library. Only 75 papers were selected for review and analysis based on exclusion criteria. The results of the systematic review have focused mostly on recent studies of Chatbots where it offers better implementation techniques for e-commerce; a bibliometric mapping of the extracted studies is also provided.
Keywords:	Chatbots, E-Commerce, Chatbot Implementation,, Systemetic Literature Review
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	IMPLEMENTATION OF ARTIFICIAL POTENTIAL FIELDS AND LYAPUNOV STABILITY AND CONTROL IN OBSTACLES AVOIDANCE OF MOBILE ROBOT USING ROS GAZEBO
Author:	YOUSSEF MHANNI, YOUSSEF LAGMICH
Abstract:	This paper presents two intelligent methods of robotics; the artificial potential field (APF) and Lyapunov stability methods as they are both designed to ensure that the robot stays clear of immovable objects or obstacles and moves in the most effective way possible toward the target. Furthermore, to address robot path planning issues in real-time using these methods, the robot can move to the target in an optimal environment while avoiding obstacles. It can also reach the target point in a limited time and choose the best and the shortest possible path. Additionally, when it calculates the best path, the robot would be obliged to move to the chosen target as the control and stability algorithm guarantees that efficiency. Moreover, the error percentage of the Lyapunov stability method would be almost zero. Ros (Robotic Operating System) Gazebo with robot waffle_Pi type was used for the simulation results demonstration.
Keywords:	Lyapunov, Obstacle avoidance, Ros, Artificial Potential Field, Mobile robot, Gazebo, Waffle_Pi,
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	CENTRAL BANK DIGITAL CURRENCY ENGAGEMENT RESEARCH BASED ON PROJECT INDEX AND USE OF DISTRIBUTED LEDGER TECHNOLOGY TO ENABLE DIGITAL CURRENCY
Author:	PRATIBHA PANWAR, SOHIT AGARWAL
Abstract:	The current payment system is rapid but has limits in terms of internet consumption, visibility, and usability. It is based on physical cash and online transactions. Central bank digital currencies strive to tackle the difficulties that afflict the current economic system by providing total transparency of the money supply. As opposed to completely decentralized cryptocurrencies, central bank-issued digital currencies are the most centralized. However, centralization will be advantageous in terms of money programmability, improved transparency, and financial crime monitoring. Before digital currencies can be widely used, a strong security framework, data protection, and effective governance will be required. Major central banks around the world are exploring technologies and methodologies to deploy their own version of digital currency that can coexist with the current physical cash and in near future replace the traditional form of value transfer. It can be promising for enhancing the economic outreach of a country and releasing a broad range of possibilities when it comes to financial inclusion and ease of transactions. CBDCs will not only help in easing payments but will help in government fund disbursements directly to eligible individuals without the interference of an intermediary bank, frictionless and faster value transfer, and exploring new financial mechanisms. These systems to support the CBDC architecture will require a robust security system that can be used for secure payments but the system design should also be simple and flexible enough to accommodate future design changes.
Keywords:	Central Bank Digital Currency, Project Index, Distributed Ledger Technology, Zero Trust, Digital Currency.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	INTELLIGENT TUTORING SYSTEMS IN EDUCATION: A SYSTEMATIC REVIEW OF USAGE, TOOLS, EFFECTS AND EVALUATION
Author:	HAZEM A. ALRAKHAWI, NURULLIZAM JAMIAT, SAMY S. ABU-NASER
Abstract:	Computer learning has grown more integrated with artificial intelligence approaches as technology has advanced, allowing for the development of more customized educational systems. Intelligent Tutoring Systems (ITS) are the name given to these systems (ITSs). This article concentrated on the many applications of ITSs created in various educational disciplines. The original studies were gathered from the Scopus, Web of Science, and Google Scholar databases between 2016 and 2022. Finally, depending on inclusion criteria, 36 publications were included in the research. Computer Sciences (36.1%) were the most common educational fields in ITSs. And Social Science, Medicine, Engineering, and Mathematics, with 27.7%, 13.8%, 8.3%, and 5.5% frequency, respectively were the most often used artificial intelligence approaches in ITSs. ITSs may use these strategies to provide adaptive guidance and training, assess learners, establish and update the learner's model, and categorize or cluster learners. The PRISMA technique is employed in this systematic literature review to select the related research and to discuss the many usage of ITSs, the tools used, the effects, and evaluation methods. Finally, we discuss the significance of ITSs in the educational process and their influence on student accomplishment, since they were utilized for school and university students and had a substantial impact on their level growth.
Keywords:	Intelligent Tutoring System, Intelligent Tutor System, Intelligent Tutoring Application, Intelligent Tutor Application, ITS, ITSB, Education
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	PREDICTION OF THE ACT OF TERMINATION OF CAR INSURANCE CONTRACTS AT THE END OF THEIR TERM
Author:	ROUAINE ZAKARIA, AMRANI AYOUB
Abstract:	The success of organizations depends not only on retaining customers but also on preventing their defection. However, there is little research on the termination of the company's and its customers' relationship. In other words, a better understanding of the factors involved in the termination process will make it easier to prevent and avoid termination while trying to recover lost customers and attract new prospects. Through this paper, we have attempted to detect, through a literature review, the "determinants of termination" that hurt the relationship between companies and their customer portfolio, and to quantify the impact of each determinant on termination behavior using an extension of generalized linear models, namely the binary logistic regression method.
Keywords:	Car Insurance, Termination Behavior, Generalized Linear Models, Binary Logistic Regression
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	SPEECH EMOTION RECOGNITION SYSTEM PERFORMANCE ANALYSIS WITH OPTIMIZED FEATURES USING DIFFERENT CLASSIFICATION ALGORITHMS
Author:	KOGILA RAGHU, MANCHALA SADANANDAM
Abstract:	Affective computing is becoming increasingly significant in the interaction between humans and machines. Emotion recognition of spoken language is a hot topic in Human Computer Interaction (HCI). Understanding a person's physical and mental state could be greatly aided by learning to identify the emotions conveyed in their speech. There are a number of practical uses for emotion recognition from Speech and it has much interest research domain in recent years. Many of the current options, however, are not yet suitable for use in production environments. The system is divided into three phases: features extraction, features selection/dimensionality reduction and classification. The first step involves extracting a wide variety of features, including prosodic and spectral components, Speech and glottal-waveform signals are used to construct long-term statistics. To tell apart associated emotions is a major challenge for SER systems. These features improve the SER's capacity to distinguish between emotions in speech. Inevitably, this high-dimensional feature vector will contain some repetition. In the second phase, the dimensionality of the feature vectors is reduced through the application of feature selection method such as Auto-Encoder technique described by the authors. Next, several classifiers, including K-Nearest Neighbour (K-NN), Logistic Regression (LR) , Support Vector Machines (SVM), Multi-Layer Perceptrons (MLP) and Convolutional Neural Networks (CNNs) are used to the optimize feature vector in stage three. Two widely-cited datasets serve as the basis for experimental evaluations of method efficacy. The Database for Emotions in Telugu Language (DETL) of a native Telugu Language and Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) of English Language are two such collections. Experimentation is carried out with the different Feature Extraction methods such as MFCC, MFCC+∆MFCC+∆∆MFCC, MFCC+∆MFCC+∆∆MFCC+Pitch+ZCR (41-dimension) and Optimized Features (30-dimension) along with the different Classification Methods. From the results, it is observed that the suggested method has potential for use in speech emotion recognition. Furthermore, when compared to other methods in the context, our approach is clearly superior in terms of classification performance accuracy rate. Here CNN model got highest accuracy with Optimized Features for RAVDESS 95.42% and for DETL 98.37%.
Keywords:	Affective Computing, HCI, Features, K-NN, LR, SVM, MLP, CNN, RAVDESS, DETL.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	SENTIMENT ANALYSIS BASED TWITTER TWEETS CLASSIFICATION USING DATA EMBEDDED WITH LSTM TECHNIQUE
Author:	SHIRAMSHETTY GOUTHAMI, DR . NAGARATNA P . HEGDE
Abstract:	Over the last two decades, social media sites have established themselves into our regular lifestyle. Collecting information from social media, following trends in social media, and knowing about people's feelings and emotions on social media are all very important today. Twitter sentiment analysis is an application of sentiment analysis on data from Twitter (tweets), in order to extract sentiments conveyed by the user. Twitter emotion recognition has gained a lot of attention nowadays due to its numerous uses in the business and government sectors. It is possible to conduct many analyses using this source. One of the most essential of these analytics, sentiment analysis, is gaining popularity. Real-time messaging and opinion sharing in social media websites have made them valuable sources of different kinds of information. Sentiment analysis attempts to extract beliefs, attitudes, and feelings from social media platforms like twitter. On the other hand, Deep Learning (DL) approaches have been increasingly popular among researchers in recent years and offer solutions to a wide range issues. However, there is a need to explore the efficacy of real-time systems that includes popular and real-time focused tweets. Hence in this approach, Sentiment analysis based twitter tweets classification using data embedded with LSTM technique. The LSTM networks and convolutional Neural Networks have shown to be effective for sentiment analysis applications. In this analysis, Long Short-Term Memory (LSTM) method is used to evaluate a tweet's emotional content. This presented approach will achieve better results in terms of Accuracy, Precision and True Positive Rate (TPR).
Keywords:	Sentiment Analysis, Deep Learning, Twitter, Long Short-Term Memory, Valence Aware Dictionary for sEntiment Reasoner (VADER).
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	A SYSTEMATIC LITERATURE REVIEW OF MACHINE AND DEEP LEARNING-BASED DETECTION AND CLASSIFICATION METHODS FOR DISEASES RELATED TO THE RESPIRATORY SYSTEM
Author:	QASEM M. M. ZARANDAH, SALWANI MOHD DAUD, SAMY S. ABU-NASER
Abstract:	Deep Learning (DL) is a sub field of Machine Learning (ML) that has considerable prospective in many areas of study like computer vision, image, and audio processing. A great number of studies using DL methods are published yearly. The emphasis of this study is on diseases related to the respiratory system detection and classification using audio files. This study outlines a systematic literature review that harmonizes the indication related to diseases related to the respiratory system detection and identification published during the years between 2015 and 2021. The goal is to collect, analyze, and epitomize the indication related to DL identification, classification and detection of diseases related to the respiratory system. The importance is that the review will aid professionals and specialists to comprehend how DL methods can be used in this respect and theoretically further backing more precise detection of diseases related to the respiratory system. 47 articles are analyzed in the current review. Most of the articles published emphasis on supervised learning with deep convolutional neural networks. Moreover, a great number of articles use data sets that are privately-prepared as opposed to open source data sets. The outcomes likewise show a shortage of tools, that is an obstruction in adjusting academic research in industrialized circumstances.
Keywords:	Literature Review; Deep Learning; Respiratory System; Classification; Machine Learning; Detection; Audio
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	RESILIENT ARTIFICIAL FISH SWARM OPTIMIZATION-BASED ENHANCED CONVOLUTIONAL NEURAL NETWORK FOR AUTISM SPECTRUM DISORDER CLASSIFICATION
Author:	B. SURESH KUMAR, D. JAYARAJ
Abstract:	Autism Spectrum Disorder (ASD) is on the rise, making early diagnosis and intervention crucial for those living with the illness. The functional connectivity deficits in ASD have been employed with neuroimaging approaches to describe complicated biomarkers. ASD is still often diagnosed using a symptom checklist gleaned through in-person evaluations. Existing computer approaches sometimes produce inaccurate diagnostic categorization when applied to massive aggregated data sets. This paper proposes a novel bio-inspired optimization-based classification model combining convolution neural network and fish swarm optimization, namely Resilient Artificial Fish Swarm Optimization-based Enhanced Convolutional Neural Network (RAFSO-ECNN). RAFSO-ECNNis powered with three optimized behaviors derived from the natural characteristics of fishes. The performance of ECNN Layers is strengthened with RAFSO. The performance of the RAFSO-ECNN is evaluated with the Autism Brain Imaging Data Exchange II (ABIDE-II) dataset, a global multisite collection of functional and structural brain imaging data. The classification accuracy and the f-measure of RAFSO-ECNN are much higher than those of the state-of-the-art classifiers.
Keywords:	Autism, Classification, Optimization, Convolutional Neural Network, Fish Swarm
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	PARTICIPATIVE DECISION-MAKING IN TERRITORIAL INTELLIGENCE
Author:	YOUSSEF AMRAOUI, SOULHI AZIZ
Abstract:	In territorial cooperation or partnership context, the participation of the territorial actors and experts in public action can generally only be calculated by observing the geographical, anthropological, historical, economic, sociological and political stakes of territorial action. This research work proposes an improvement of our models of participatory decision-making in an environment of territorial cooperation or territorial partnership [1]. The main improvement of our model and the enhancement of the expertise and experience of actors and territorial experts, given the extreme importance of the latter in the regulation, orientation and negotiation of territorial action, this improvement based on fuzzy set theory, which used to solve complex and uncertain system problems.
Keywords:	Fuzzy Logic, Cooperation, Participation, Territorial Intelligence, Indicators, Decision-Making.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	CARDIOVASCULAR ABNORMALITIES DETECTION THROUGH IRIS USING THRESHOLDING ALGORITHM
Author:	NILAM UPASANI, ASMITA MANNA, SHATABDI PINGALE, YASHASHREE SHINDE, SAKSHI RATHI, SONALI SURPATNE
Abstract:	Cardiovascular diseases have proven to be the leading reason of death worldwide. To identify cardiovascular diseases at an early stage, often very expensive pathological tests are required. A less costly alternative method for determining the conditions of the organs is highly appreciated, and Iridology is one such popular method. Many researchers have proposed cardiovascular disease identification systems by combining Iridology with computation system. In this study, a novel model for detecting heart abnormalities using Iridology is proposed. The entire process includes several stages such as target capture, pre-processing, auto-cropping based on histogram analysis, heart area extraction, and classification using a thresholding algorithm. The cropping method and classification process are both affected by the iris photography procedure. The iridology expert at Clinic in Indonesia labelled the data as abnormal and normal. The precision produced by the system ranges from 80-83%. Some errors occur due to ineffective cropping. The failed outcome may affect the segmentation process resulting in erroneous segmentation in the heart area eventually.
Keywords:	Iridology, Pre- processing, Classification, Thresholding Algorithm, Binarization.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	A HYBRID APPROACH FOR REQUIREMENTS PRIORITIZATION IN AN INCREMENTAL DEVELOPMENT MODEL
Author:	FATIMA THAHER ABUROMMAN
Abstract:	In incremental development model (IDM) stakeholders are involved in the development process to add more emphasis on the system implementation phases rather than the requirements analysis and system design, which in turn may give better results in terms of the delivery of the components of the system which are referred to as increments. Because the development process is shared with stakeholders experience, interest, positions and other factors, it might not be easy to rely on the stakeholders opinions on deciding which increment will be the next. This paper provides a solution to this problem by forming a mathematical representation and model that is referred to as the Hybrid Approach (HA) which is a hybridization between the dept-first search (DFS), the value-oriented prioritization (VOP), and the greedy algorithm. The HA model is compared to the analytic hierarchy process (AHP) approach. The results show that the HA model outperforms the AHP approach in terms of runtime.
Keywords:	Increment Prioritization; Incremental Model; Software Engineering; Requirements Prioritization; Hybrid Model; Value-Oriented Prioritization.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	IMPORTANT FACTORS THAT AFFECT CUSTOMER SATISFACTION WITH DIGITAL BANKS IN INDONESIA
Author:	YOS VINCENZO, RIYANTO JAYADI
Abstract:	Customer satisfaction has been considered as the measure of information system investment success in many businesses. Customer satisfaction could be difficult to clearly define but is considered as crucial evaluation construct for business investments. Covid-19 has triggered many financial institutions to invest heavily in technology to improve customer satisfaction and also generating more interaction. Indonesia banking industry evolution and revolution happened in accelerated manner to address this need. Traditional banks are creating and launching their digital applications, new banks are launched as digital bank. These banks invested significantly in building these digital solutions. In order to be successful, important factors influencing customer satisfaction in using the application should be considered to continuously improved the digital application and sustaining the business of this digital banks. This research aimed to evaluate hypothesis related to customer satisfaction in using these digital banks application by factoring in Ecosystem, Company Image, Promotion, Perceived Usefulness and Actual System Use. The benefit for this research was aimed to provide insights for digital banks to improve their strategy for their digital application and in the end will benefit the business. The result from the research contributed that ecosystem definitely a key component that need to be further researched to increase customer satisfaction with the digital bank application while also there is a need to do a deep dive for the offers/capability in the digital bank apps addressing specific customer needs to ensure they stay and use the digital bank application and less switching to other provider/banks; and in the end increased satisfaction and business for the bank.
Keywords:	Digital Bank, Customer Satisfaction, Ecosystem, Promotion, Indonesia
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	A SYSTEMATIC LITERATURE REVIEW OF DEEP AND MACHINE LEARNING ALGORITHMS IN CARDIOVASCULAR DISEASES DIAGNOSIS
Author:	ZAKARIA K. D. ALKAYYALI, SYAHRIL ANUAR BIN IDRIS, SAMY S. ABU-NASER
Abstract:	During the whole cardiac cycle, heart sounds are created, and blood enters the heart chambers as the cardiac regulators open and close. Blood flow produces aural noises; the more turbulent the blood flow, the more ambiances are created. Two common cardiac sounds occur in sequence with each heartbeat in healthy adults. These are the first heart sound (S1) and second heart sound (S2), which are caused by the closure of the atrioventricular and semilunar valves, respectively. The current systematic review depends on “the Preferred Reporting Items for Systematic reviews and Meta-Analysis statement” and 40 appropriate studies. The search of the literature employed search engines similar to: IEEE Xplore, Google Scholar, Hindawi, PubMed, SCOPUS, Wiley Online, Web of Science, Taylor and Francis, ScienceDirect, and Ebscohost. This study concentrated on four characteristics: Algorithms of Machine and Deep Learning, best-algorithm performance, datasets, and application used in Cardiovascular Diseases predictions. The experimental articles did not use Reinforcement Learning, Semi-supervised learning, promising aspects of Deep and Machine Learning. Algorithms based on ensemble technique exhibited sensible rates of accuracy nonetheless were not frequent, whereas Convolutional Neural Network (CNN) were well epitomized. A few studies smeared main datasets (13 of 37). Recurrent Neural Network (RNN), boosting algorithms, Support Vector Machine (SVM) and K-Nearest Neighbors (KNN), were the best performing algorithms. This review will be beneficial for investigators predicting Cardiovascular Diseases using machine and deep learning methods.
Keywords:	Cardiovascular Diseases, Datasets, Algorithms, Deep Learning, Machine Learning
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	EFFECTS OF E-CUSTOMER SATISFACTION AND E-CUSTOMER LOYALTY AND GUANXI MODERATION FOR ELECTRONIC SHOPPING CHANNEL FASHION PRODUCTS
Author:	SCHERLY HANSOPAHELUWAKAN, MTS ARIEF, ELIDJEN, VIANY UTAMI TJHIN
Abstract:	The purpose of this study is to find out whether E-Customer Satisfaction has a significant effect on E-Customer Loyalty moderated by Guanxi in this pandemic era, and then to find out the factors that influence it. To understand how customer satisfaction is improved through omni-channel retail, we collect data through survey research, specifically surveys. Research data will be collected using an online survey involving e-customers who have omni-channel experience in fashion during the pandemic. This quantitative research will be analyzed using the Structural Equation Model. In order to produce concise and accurate findings, the data is analyzed to investigate the reasons behind the preferred shopping methods and what influences those purchases. The results of the analysis show that Guanxi has an effect on e-customer loyalty. In addition, the omni-channel customer experience and e-customer satisfaction simultaneously influence e-customer loyalty. Finally, this study creates opportunities for future research and outlines potential insights for increasing customer satisfaction and loyalty. The novelty of this research is that this research is a research on the concept of e-Customer Loyalty which is moderated by e-Customer Satisfaction which is mediated by the guanxi Concept
Keywords:	Customer Satisfaction, E-Customer Satisfaction, Omni-Channel Customer Experience, E-Customer Loyalty
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	ACADEMICS BEHAVIORAL INTENTION AND USAGE OF IOT IN E-LEARNING: MODERATION OF GENDER AND EXPERIENCE
Author:	FAIQ AZIZ, AZIZI SAFIAI , NOR WAHIZA ABDUL WAHAT , SITI RABA AH HAMZAH , SEYEDALI AHRARI , NOMAHAZA MAHADI
Abstract:	As the world shifts to e-learning, IoT is becoming increasingly important in the learning and teaching environment. The key concerns are the elements that influence academics behavioral intention to adopt IoT, and how the whole operation affects their performance. However, there is a research gap in past studies that have not addressed this issue sufficiently. As a result, this study employed the Unified Theory of Acceptance and Use of Technology (UTAUT) as a guideline to examine the factors that influence academics behavioral intentions to use IoT. Furthermore, the moderating effects of both gender and level of experience on this relationship were inspected. The structural models were validated, and the predefined hypotheses were presented (n = 321). The results from the Structural Equation Modeling approach using Amos 26 indicate that performance expectancy, social influence, and effort expectancy directly influenced behavioral intentions to utilize IoT. The findings also showed that facilitating conditions were the most important determinant of academics actual usage of IoT. The structural model was further investigated according to the experiences of the male and female academic groups. The findings revealed a different pattern of strength and significant relationships between groups with the overall model, implying that gender and experience act as moderators. This study provides a wealth of antecedents from which to construct a thorough theory of IoT adoption. The theory explores the elements that influence academics willingness to utilize IoT from the standpoints of the technology itself, social context, and individual user characteristics. By employing the proposed approach, Universities can modify their EL strategies to make the most of their resources and in turn improve efficiency.
Keywords:	IoT adoption; UTAUT; Academics; E-Learning; Gender; Experience
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	WEB OF THINGS ENABLED SMART CAMPUS IN SUPPORTING LEARNING
Author:	ERIN, TOGAR ALAM NAPITUPULU
Abstract:	Today the development of technology is growing rapidly. One of the fields that participate in technological developments is the field of education, especially universities. Technological developments help universities become smart campuses. Smart campuses require the use of technologies such as information systems, Internet of Things (IoT), Artificial Intelligence (AI) technology and so on. However, the focus of the problem in this paper is the utilization of IoT. IoT has a problem, namely the problem of interoperability. This interoperability problem can be solved using web resources, also known as the Web of Things (WoT). In this paper, we will discuss how to build a prototype WoT system where everything can be accessed via the web to solve IoT interoperability problems. The design that will be carried out in this writing to build WoT starts from controlling physical things, accessing data, and connecting things to things (web to web). The prototype development of this system will use the Raspberry Pi and Node-RED. The result of this prototype is that it can access lights and access data from a web, process it, and display it in graphic form by connecting it to Google charts. The results of this light control prototype can be a model that can be learned to control things on campus so that it can create a smart campus. The results of communication between things and things can also be used to integrate all data on campus so that it can be used easily. For further work it is necessary to evaluate to get feedback about the system and to find out that the system is running well and can help users on campus. Models or prototypes can be easily extended or scaled to wide implementation aspects of smart campuses, including ITS embedded in personalized learning facilitation.
Keywords:	Smart Campus, Internet of Things, Web of Things, Node-RED, Raspberry Pi
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	SUCCESSION PLANNING TECHNOLOGY TRENDS AND THE INFLUENCING FACTORS: A SYSTEMATIC LITERATURE REVIEW
Author:	DEVYANO LUHUKAY, FORD LUMBAN GAOL, MEYLIANA, HARJANTO PRABOWO
Abstract:	Growth in Organization can trigger a change in the organizational structure that could happen suddenly, by design, or as predictable. Such as promotion, resignation, or any other process that could cause the organization's leadership emptiness. Seeing this as an essential thing, the organization must have succession planning. Succession planning can prepare successors systematically and continuously. Conventional succession planning has been implemented in many organizations but without involving technology in all the processes. Machine Learning techniques can be used to improve succession planning. To see whats the trends of succession planning articles and to know what factors influenced succession planning specifically in the candidate selection process, this paper used a Systematic Literature Review from Kitchenham as the research method. The database of this study gathers information from the early Covid19 pandemic era to see how succession planning was before the impact of the Covid19 pandemic. This paper found three trends in succession planning focus areas: concept, research, and thesis or dissertation product. Besides that, there are 23 factors extracted from the papers that influenced succession planning. The implementation of methodology such as machine learning is also discussed in the result papers to improve the process and implementation of Succession Planning in the organization.
Keywords:	Succession Planning, Machine Learning, Human Resources, Education, Systematic Literature Review
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	EVALUATION OF BLOCKCHAIN TECHNOLOGY ACCEPTANCE FACTORS IN THE TOKOCRYPTO APPLICATION
Author:	SHEREN PRISSCILYA, TOGAR ALAM NAPITUPULU
Abstract:	The development of the internet which is growing rapidly at this time has given rise to various recent technologies in the world, one of which is cryptocurrency. The total level of cryptocurrency ownership globally averages 3.9 percent out of 300 million crypto users in the world. Additionally, an estimated 18,000 businesses have opened to accept cryptocurrency as payment. However, the increasing popularity of cryptocurrencies also raises questions about financial risks in various countries including Indonesia. The Indonesian Financial Services Authority (OJK) prohibits financial service institutions ranging from banking, and insurance, to finance to facilitating cryptocurrencies. Although not as a means of payment. However, cryptocurrencies can be included as commodities that can be used only on futures exchanges and the public can buy cryptocurrency assets through any trade registered with BAPPEBTI. Tokocrypto is the first crypto asset trader listed with BAPPEBTI. However, the Tokocrypto application rating is still considered unsatisfactory. Based on Ratings and Reviews on the Tokocrypto application, the lack of satisfaction in investing in cryptocurrency assets through the Tokocrypto application is caused by the inadequate quality of the application. The purpose of this research is to find out and identify what factors are one of the reasons for user acceptance in investing in cryptocurrency assets through the Tokocrypto application. The research was conducted by distributing questionnaires to 398 respondents and processing the data that had been collected by analyzing the data using the PLS-SEM method and getting the results that cyber security, habit, trust, social influence did not have a significant impact on intention to use. On the other hand, the influence of government regulations, information quality, and system quality on intention to use, and intention to use on usage behavior has a significant positive impact.
Keywords:	Blockchain, Cryptocurrency, SmartPLS
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	A HYBRID DEEP CNN-LSTM NETWORK (HDCLN) FOR SHOULDER IMPLANT CLASSIFICATION USING X-RAY IMAGES
Author:	RAJENDRA PRASAD BANAVATHU, JAMES STEPHEN M, PRASAD REDDY P.V.G.D
Abstract:	Today, prosthesis manufactured from polyethelene and metallicconstituents have been widely used to replace the impaired ball and sockets of the human shoulder. Years after the replacement, there may be need for reoperation and revision due todesecration in the quality of the prosthesis. This process requires the information about the prototype as well as the corresponding manufacturer of the prosthesis. In some situations, the patient and the primary doctor may not have the information about the prototype and manufacturer of the prosthesis. Usually, manual recognition of the prototype and manufacturer of the prosthesis is carried out during the preoperative planning. However, the manual identification of the model and manufacturer is time-consuming and prone to error. An automatic model identification and manufacturer classification system can speed up the treatment process and reduce the operation risk associated with the manual identification system.In this paper, we introduce a hybrid deep CNN-LSTM network for identification and classification of implant manufacturer. In this network, we used the CNN layer to extract features from the radiographic images and the LSTM layer for identification and classification of implants. A collection of 597 implant images from four manufacturers, including 83 images from Cofield, 294 from Depuy, 71 from Tornier, and 149 from Zimmer, were used as dataset to train and test the model. Based on experimental results, our model achieved the accuracy of 98.7%, precision of 98%, F1-score of 98.2% and recall of 98%. Based on the performance of the model, we believe that this model will be a useful tool in preoperative planning and can be applied in the identification and classification of implants from other manufacturers.
Keywords:	CNN-LSTM, Manufacturer, Classification, Shoulder Arthroplasty
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	THE IMPACT OF INFLUENCERS REPUTATION AND MENTAL SIMULATION ON BRAND EVALUATION
Author:	YISITIE XING, JING-YUN ZENG, CHANG-HYUN JIN
Abstract:	This study explores how the reputation of Internet influencers affects consumers evaluation of brands and purchase intention. Furthermore, it examines whether consumers brand evaluation or intention depends on the type of mental simulation and level of need for cognition. The ripple effects of the internet influencers reputations were identified by constructing a 2x2x2 experimental design factoring the degree of influencer reputation, mental simulation, and need for cognition. The results revealed that the brand evaluation or purchase intention differed depending on the degree of influencer reputation and type of mental simulation. Consumers believe that influencers reputations are an important factor to evaluate a product or brand. In process simulation, brand evaluation and purchase were also higher. For outcome simulations, the product may have been evaluated focusing on realistic benefits or direct desires. Consumers who enjoy cognitive efforts make careful decisions when evaluating brands and demonstrate higher willingness to purchase. The two-way interaction between mental simulation and need for cognition has a significant impact on brand evaluation and purchase intention. Additionally, the differences in psychological simulation effects of consumers” internal factors when evaluating or purchasing new products are studied. The study has practical importance for marketing professionals as it establishes the impact of individual need for cognition on the simulation effect.
Keywords:	Internet Influencer, Reputation, Mental Simulation, Need for Cognition, Brand Evaluation, Purchase Intention
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	AN EFFECTIVE HYBRID FEATURES FOR DRIVER FATIGUE DETECTION USING CONVOLUTIONAL NEURAL NETWORK
Author:	V.VIJAYPRIYA, M. UMA
Abstract:	Drowsy driving is one of the leading causes of road accidents worldwide. Drowsiness is caused by a lack of sleep and irregular sleeping patterns. People who are drowsy may have difficulty driving and cause major accidents, as evidenced by the high number of injuries and deaths worldwide. As a result, detecting driver drowsiness is critical in order to save many lives. The proposed model made use of non-intrusive methods related to fatigue states. The proposed system employs a multi-task convolutional neural network to predict driver fatigue states based on the percentage of eye closure (PERCLOS) and yawning detection in a video stream. Faces are identified using facial landmarks in the extracted images from the video streams, and the trained model predicts whether the person's face in the image is fatigued with closed eyes and yawning or non-fatigued with open eyes, talking, and a closed mouth. In comparison to the existing model, our proposed model has been tested on the D3S publicly available datasets and has produced better classification models for predicting fatigue states with 97% accuracy using deep learning models.
Keywords:	Convolutional Neural Network, Driver Drowsiness, Facial Landmark, Percentage Of Eye Closure
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	A FRAMEWORK FOR DEVELOPING SECURE INTERNET OF MEDICAL THINGS: A COMPREHENSIVE ROADMAP FROM AN ARTIFICIAL INTELLIGENCE PERSPECTIVE
Author:	BANDAR M. ALSHAMMARI
Abstract:	The Internet of Medical Things (IoMT) has gained an extensive reputation in many applications within medical sectors. The late expansion of using these technologies has helped many organizations to progress well by delivering high-quality services. Such advancements and developments wouldnt hide the fact that such technologies also have many risks associated with their usage. Such risks vary in their severity depending on many factors, including the type of services IoMT provides and the customers it serves. However, the most important risks these systems are lately facing are related to their security. Therefore, any IoMT has to be secure enough in order to gain the trust and credibility of its clients. Although the IoMT security issues have been considered in several studies, enough attention has not been given to the best practices for developing secure applications. In fact, not following a precise and well-structured framework for developing secure IoMT applications has lately resulted in many cyber incidents directed toward such IoMT applications. The novelty of this work is to develop a framework that resolves such issues from the early stages of development. The main purpose of this framework is to provide a roadmap consisting of several strategic approaches for developing secure IoMT applications that take into consideration the advancements in artificial intelligence. The framework will also guide IoMT developers on how to mitigate and detect vulnerabilities, and hence the cyberattacks associated with each layer of a specific IoMT, using relevant machine/deep learning methods.
Keywords:	Cybersecurity; Security Attacks; Artificial Intelligence; Internet of Medical Things (IoMT); Big Data.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	ARTIFICIAL IMMUNE SYSTEM: A SYSTEMATIC LITERATURE REVIEW
Author:	NIDA HASIB, SYED WAJAHAT ABBAS RIZVI, VINODANI KATIYAR
Abstract:	Purpose: This work seeks to make clear that Artificial Immune Systems are computational paradigms that have drawn researchers to develop immune based models and methodologies using theories, processes, and algorithms in many domain areas to tackle various social or technical difficulties. Additionally included are the AIS methods and algorithms that have been applied during the past 20 years to deal with clustering, classification, and optimization problems. Design/ Methodology/Approach – An analysis of the existing literature on artificial immune systems was compiled by the authors as part of their systematic literature review, which also served to highlight any gaps in the body of knowledge and set the stage for future study. Findings: The evaluation process indicates the pervasive influence of Artificial Immune System approaches employed in issue domains other than software development projects but elevated the necessity of AIS has been examined to classify modules of software development projects. Algorithms are built on Immune system theories and computational features used in different domains to achieve the best classification accuracy. The authors found out that, compared to the number of proposed approaches, implementation of software project development approaches is lacking. The study further reveals that such approaches lack completeness from one or more perspectives. Originality/Value: Since there are numerous experience reports and case studies on Artificial Immune Systems in the research literature, as far as the authors are aware, there is no systematic overview and synthesis of this expanding field of study. The authors organize the review process on the Artificial Immune system and their studies in various domains between 2000-2021. We used a systematic process to select one hundred four articles, twenty-eight conference proceedings, seven book sections, four workshop proceedings, eight symposiums, two technical reports, and five theses. This study focuses on various domains and their proposed approaches built on Artificial Immune Systems in the last two decades.
Keywords:	Artificial Immune System, AIS, Systematic Literature Review, SLR, Biological Immune System, BIS
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	RELIABLE PARAMETRIC FILTERING FOR VIDEO SYSTEM NOISE IMMUNITY
Author:	SAFIN R.T., AMREEV M.B., PAVLOVA T, GARMASHOVA Y, SATIMOVA Y, SERIKOV T.G.
Abstract:	The article discusses issues of robust signal filtering by a newly developed method devoid of the following disadvantages of known systems, namely: - working only with a fixed spectrum of the input signal; - inability to operate in signal processing systems with a changing spectrum; - narrow band. The filtering method is based on: - elimination of operation at a single fixed frequency and the possibility of working in systems with a changing spectrum is achieved by using a clock pulse generator, the frequency of which varies depending on the width of the input signal spectrum due to the use of frequency-dependent coefficients; - elimination of the lack of narrowband, which is achieved by using a signal spectrum width analyzer and a nonlinear variable filter capacitance (varicap) depending on the width of the signal spectrum. The substantiation of the proposed method is given, which allows processing signals with varying spectrum width, increasing their noise immunity and reducing the level of pulse interference, including in video surveillance and television systems. Modeling and experimental studies have been carried out to confirm the operability of this method.
Keywords:	Robust Filtering; Bandwidth Changing; Nonlinear Filter; Video Pulse
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	FINANCIAL TECHNOLOGY AND EMOTIONAL INTELLIGENCE ON FINANCIAL MANAGEMENT BEHAVIOR OF FEMALE MSME OWNERS
Author:	NADIA ASANDIMITRA, DEWIE TRI WIJAYATI, ACHMAD KAUTSAR, AND ACHMAD MURDIONO
Abstract:	East Java's economy relies heavily on the contributions of its women-folk. The mediating role of locus of control in the relationship between financial knowledge and management behavior is still being investigated. This study employs a purposive sampling strategy to collect data from a population of female MSME owners in East Java, utilizing a causal explanatory research design. In this research, SmartPLS 3.0 was utilized for data analysis. The study found that while financial knowledge and financial technology did not influence financial management behavior, financial attitude, financial self-efficacy, financial literacy, and emotional intelligence did. However, other findings show that the link between financial literacy and prudent decision-making is not mediated by locus of control.
Keywords:	Emotional Intelligence, Financial Knowledge, Financial Management Behavior, Financial Technology, Locus of Control.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	GRAPH DATABASE AND RELATIONAL DATABASE IN DETECTING MONEY LAUNDERING CASES: A LITERATURE REVIEW SYSTEM
Author:	DHIKA NUR AISYAH, FAUZAN RISQULLAH, EDWIN TUNGGORO, INDRAJANI SUTEDJA
Abstract:	Money laundering is the illegal process of making money from criminal activities, such as drug trafficking or terrorist financing, manipulated so that the money appears to come from a legal source. Due to the secretive nature of money laundering, it is difficult to accurately estimate the total money that goes through the money laundering cycle. However, it is estimated that the amount of money laundering in 1 year globally reaches 2-5% of global GDP, or 800 billion - 2 trillion US dollars. The money laundering detection system is currently built using a relational database system, where the system has limitations in terms of performance so that tracking financial transactions in money laundering cases is very difficult. The limitations of this relational database system can be mitigated with the capabilities of the graph database. This is due to the nature of the graph database itself, which uses graphs that represent entities as nodes and relationships between entities as edges. The main advantage of a graph database is that the relations are stored together with data that can be obtained with just one query. Therefore, the purpose of this paper is to determine the performance of graph databases when compared to relational databases in detecting money laundering cases. The method used is by using a System Literature Review of 27 relevant papers. The final result obtained is a graph database capable of tracking the flow of funds in money laundering cases found on the transaction network. The conclusion obtained is that the graph database performance in running queries is better than relational databases, especially processing data that has many relationships.
Keywords:	Graph Database, Money Laundering, Relational Database
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	RESTAURANT RECOMMENDATION SYSTEM USING ADVANCED COLLABORATIVE FILTERING AND REVIEW TEXT CONTENT APPROACH
Author:	DANIEL RIANDY, AMALIA ZAHRA
Abstract:	Recommendation system can provide recommendations to consumers on which restaurants might be liked by these consumers. This method plays the major role in food industries. The approach taken is to use collaborative filtering for the final touch in providing restaurant recommendations, but before this stage, it will go through several stages, including Natural Language Processing (NLP) aiming to create a new scoring which is the result of sentiment analysis combined with previous ratings (review text content). Then, clustering will be carried out on customers, to accelerate and optimize the recommendation process. From these two approaches, collaborative filtering is then carried out, so that it can provide some recommendations that are most suitable for consumers. However, in the recommendation system, there is a problem called the Cold Start Problem, this is a condition where the consumer is new so we cannot give recommendations. To overcome this problem, Location Based Filtering will be used, in this method, the consumer must fill in the location where he is located, so that we can provide recommendations for the best restaurant close to the location where the consumer is located. To measure model performance Root Mean Squared Error (RMSE) is used, the results obtained for the RMSE of the proposed method are 0.55, better than the baseline RMSE at 0.62. The time needed to provide recommendations on the proposed method is 247 ms, faster than the baseline model with a speed of 1.83 s. To overcome the Cold-Start Problem, Location Based Filtering is used, in this method the consumer must fill in the location where he is, so that he can provide the best restaurant recommendations that are close to the location where the consumer is.
Keywords:	Recommendation system, Collaborative filtering, Natural Language Processing, Machine Learning, Clustering, Cold Start Problem, Review Text Content
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	MANAGEMENT OF THE GLOBAL COMPETITIVENESS OF COMPANIES IN THE FIELD OF ELECTRONIC COMMERCE IN THE CONDITIONS OF DIGITALIZATION
Author:	SERHII AREFIEV, IRYNA SHEVCHENKO, ULYANA SAVKIV, DMYTRO HOVSIEIEV, YURII TSIZHMA
Abstract:	The study is devoted to the issue of managing the competitiveness of companies in the e-commerce field, the relevance of which is determined by the spread of the latest information and communication technologies, the spread of the Internet, the development of payment systems, the increase in the digital culture of the population, which determines the attractiveness of the global e-commerce market and opens up new opportunities for ensuring competitiveness in the market. The purpose of the study is to highlight the principles of managing the global competitiveness of companies in the e-commerce field in the conditions of digitalization. To achieve the goal, a systematic approach was used, which makes it possible to comprehensively study the effectiveness of e-commerce of companies. Based on the analysis of the the dynamics of the global e-commerce market, a number of objective reasons were identified that determine the competitive positions of certain companies in the e-commerce market. Obstacles to the development of the global e-commerce market and factors that activate the e-commerce development are identified, including: a positive effect of networks, which acts as a motivator for the e-commerce development; technology development in the digital economy; increasing the economic efficiency and competitiveness of companies due to the reduction of operating costs and the expansion of the consumer market due to the use of digital technologies. It is of practical importance to improve the effectiveness assessment of the companies entering the e-commerce market, taking into account the establishment of dense communications with the company's customers in real time, creating a positive image of the company in the e-commerce market, ensuring sales stimulation through the use of advertising measures, monitoring the activities of competitors in the e-commerce market.
Keywords:	Management, Electronic Commerce, Company, Competitiveness, Digitalization, Global Economy
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	IMPROVING FACE RECOGNITION IN LOW ILLUMINATION CONDITION USING COMBINATION OF IMAGE ENHANCEMENT AND FACE RECOGNITION METHODS
Author:	OWEN JACKSON DHARMADINATA, GEDE PUTRA KUSUMA
Abstract:	Some face recognition models are constantly challenged by low illumination. However, we can address this issue by employing a low illumination image enhancement filter on the face recognition model. This study compares four low illumination image enhancement methods: MIRNet, Adaptive Gamma Correction, RetinexNet, and Retinex. SSIM evaluates these four models using the LOL dataset. The assessment findings are applied to both the ArcFace model and Facenet for face recognition, and they are evaluated with a combination of the Essex Faces96 and Essex Grimace datasets. The final model based on our suggested approaches is a combination of MIRNet and ArcFace models and MIRNet and Facenet models. MIRNet and Facenet combination perform better in higher illumination percentages, whereas MIRNet and ArcFace combination performs better in lower illumination percentages. At both 20 and 10% illumination percentage, the combined approach of low illumination image enhancement MIRNet method and Facenet model obtained Rank-1 Accuracy of 0.96 and 0.94. While the combination of MIRNet and ArcFace achieved a 0.84 Rank-1 Accuracy for a 05% illumination percentage.
Keywords:	Face Recognition, Low Illumination Face Recognition, Deep Learning, Facenet Model, ArcFace Model
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	REVIVED ANT COLONY OPTIMIZATION-BASED ADABOOST ALGORITHM FOR HEART DISEASE AND DIABETES (HDD) PREDICTION
Author:	S. USHA, Dr. S. KANCHANA
Abstract:	Hidden information in the massive medical data collected by the healthcare industries assists in making productive decisions. Using cutting-edge data mining tools, we can get accurate findings and make informed decisions based on that information. Affordable quality treatment is a big concern for healthcare institutions like hospitals and clinics. The worlds leading killers are heart disease and diabetes (HDD), which are often misdiagnosed. Earlier diagnosing of HDD saves the victims life and the costs associated with treating the disease. Additionally, most existing machine learning algorithms tend to specialize in forecasting certain diseases only. Its possible that a classifier that can reliably forecast the frequency of several. In this paper, Revived Ant Colony Optimization-based Adaboost Algorithm (RACOAA) for HDD Prediction. The Adaboost algorithm is enhanced by fixing the threshold value for predicting the HDD, and it will assist in enhancing the classification accuracy. To predict more accurately, this research applies the enhanced version of ant colony, RACO. The deposit of pheromones is optimized for better classification. This study uses the Cardiovascular Disease Dataset and the PIMA Indian Diabetes Dataset to assess the efficacy of the proposed classifier. The assessment results show that the suggested classifier achieves higher classification accuracy than the state-of-the-art classifiers.
Keywords:	Ant Colony, Adaboost, Diabetes, Heart Disease, Optimization
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	A NOVEL CRYPTOSYSTEM OF AN UPGRADED CLASSICAL CIPHER AND RSA ALGORITHM FOR A SECURE AND AN EFFICIENT ELECTRONIC VOTING SYSTEM
Author:	SRAVANI JAYANTI, K CHITTIBABU, PRAGATHI CHAGANTI, CHANDRA SEKHAR AKKAPEDDI
Abstract:	Election is a fair decision making process by an authorized group of individuals to elect a person or a party to provide them power to take further decisions for the welfare of the voted people. Traditionally, elections were conducted using Ballot-paper system which is less efficient with a threat of confusion, tampering of votes and more time-consuming. In the digital world, Electronic Voting Systems (EVS) are introduced which are less time-consuming, efficient and prevent tampering of votes. Several research works have invaded different EVS such as Mix-net Based e-voting, Homomorphic e-voting, Blockchain e-voting etc using cryptographic tools. This paper draws attention towards developing an Electronic voting model applying a modified classical cipher and RSA cryptosystem. The developed method is implemented using a C++ program. The performance of the developed E-Voting System is analyzed in terms of the security provided, time and memory requirements.
Keywords:	Cryptology; Electronic Voting; Affine-Hill Cipher; Quadratic residues; RSA
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	EFFECTS OF VIRTUAL PRIVATE SOCIAL NETWORKING IN ACADEMIC PERFORMANCE OF STUDENTS
Author:	ALAGAPPAN ANNAMALAI, RAMESH CHANDRA POONIA, SURESH SHANMUGASUNDARAM
Abstract:	A virtual private social network (VPSN) is generated automatically amongst peers using a social media app to build ties. One of the most significant repercussions of students' excessive usage of social networking sites is a decline in their academic performance. In a study of medical students, social media and the internet were shown to harm students' academic performance and classroom attentiveness. An increasing number of studies link the use of social media to poorer academic performance, such as fewer students doing their assignments and lower test scores. Students who receive specialised training in deep learning will have the superior cognitive abilities needed to succeed in today's more cognitively demanding workplaces. It teaches children to be critical thinkers, productive members of society, and active participants in a democratic society. As a perceptron used in image recognition and processing, a convolutional neural network (CNN) processes pixel data from social networks. A CNN uses multiplayer perception to lessen the processing needs of pupils. Humans and neurons make up the VPSN-CNN network, which the article explains. Neurons generate dendrites and axons to receive and transmit signals, while humans engage with long-reaching telecommunication equipment or biological communication systems. These will help remember, learn, unlearn, and relearn what has already been learned. In courses where social networking sites were utilised in addition to traditional teaching methods, most students reported feeling more socially engaged and more positive about their educational experiences. Students' and instructors' concerns regarding the educational usage of social media are addressed with recommendations for further study and practice in better performance and accuracy for students data secure and comparison with existing methods.
Keywords:	CNN, Deep learning, Social network, Students.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text

Title:	SIMULATION OF THE RAINBOW ATTACK ON THE SHA-256 HASH FUNCTION
Author:	MANANKOVA O.A., YAKUBOVA M.Z., RAKHMATULLAEV M.A., BAIKENOV A.S.
Abstract:	The value of data is growing every day. Data is a key factor in both scientific research and public administration. The development of IT technologies has led to the generation of a large amount of personal data, which has become the basis for the development of machine learning technologies and big data processing. This growing demand is bringing renewed interest in data privacy practices and processes. The study of the strength of a hashed message is of great importance in modern authentication systems. The hashing process is inextricably linked to the password system, since passwords are usually stored in the system not in clear text, but in the form of hashes. The SHA-256 hash function was chosen to model therainbow tables attack. An algorithm for constructing a rainbow table for the SHA-256 hash function in the Cryptool 2 environment is proposed. The conditions under which the use of rainbow tables will be effective are determined. This article aims to practically show the process of generating a password and rainbow tables to organize an attack on the SHA-256 hash function. Studies show that rainbow tables can reveal a password without salt faster than with salt. As the password bit increases, the decryption time increases in direct proportion.
Keywords:	Simulation, Salt, Hash Function, Attack, Rainbow Tables, SHA-256, Cryptanalysis.
Source:	Journal of Theoretical and Applied Information Technology 28^th February 2023 -- Vol. 101. No. 4-- 2023
Full Text