		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
June 2022 | Vol. 100 No.11

Title:	AN ALGORITHM BASED ON SENTIMENT ANALYSIS AND FUZZY LOGIC FOR OPINIONS MINING
Author:	GHADA KHAIRY, ABEER M. SAAD, SALEM ALKHALAF, MOHAMED A. AMASHA
Abstract:	The collection and analysis of comments is an important topic, as traditional methods are based on analyzing and collecting students feedback through a questionnaire. This paper aims to extract knowledge to support students learning processes by understanding students sentiments regarding the use of gamification quizzes in learning. The paper attempts to apply formative assessment in an innovative way, enrich the activities accompanying the educational process in higher education, and apply fuzzy logic to obtain a more accurate analysis and reflect the true perceived sentiment in the texts content. In addition, this paper proposes an algorithm for a gamification framework based on sentiment analysis and fuzzy logic (AGFSAFL). AGFSAFL consists of four main phases: specifying the proposed gamification algorithm, preprocessing students opinions about gamification, applying sentiment analysis, and fuzzy logic classifier for student satisfaction level. Furthermore, it presents a visualization of the work of AGFSAFL. In the future, this paper intends to enhance our proposed algorithm to aid people with special needs (i.e., the blind) by converting it into a dynamic interactive application to ease their understanding of the materials.
Keywords:	Gamification, Sentiment Analysis, Student Feedback Analysis, Fuzzy Logic, Algorithm
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	A ROBUST WATERMARKING SCHEME BASED ON DCT, IWT and SVD WITH OPTIMAL BLOCK
Author:	MOHAMED RADOUANE, NADIA IDRISSI ZOUGGARI, AMINE AMRAOUI, MOUNIR AMRAOUI
Abstract:	Signal processings impact on development of digital media technologies have become a hot topic. The increased of computer network and the growth of the Internet have facilitated the production and distribution of unauthorized copies of multimedia information (text, image, sound, and video). To ensure multimedia security, researchers are focusing on digital image watermarking. With this new concept, the watermark is not just hiding in an image, but it’s marked indelibly. In this paper a robust method of digital images watermarking based on combination of DCT, IWT and SVD is proposed. At first, Visual cryptography is used to encrypt the watermark image. Then DCT is applied to it and to the host image. IWT and SVD are applied on DCT coefficients of both watermark and host images. After that, the watermarking process is done by embedding singular values of watermark image to the singular values of host image. Moreover, the obtained watermarked images are subjected to different attacks to improve the robustness of the proposed scheme. Finally, the extraction process is based on watermarked image and the reverse method of embedding process to reconstruct the original watermark. The performance is evaluated under various attacks and experimental results show that our algorithm provides a high level of robustness and imperceptibility than the state-of-the-art methods.
Keywords:	DCT, IWT, SVD, Entropy, Watermarking.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	UNVEIL ARCHITECTURAL DISTORTION WITH EMENDATED IMAGE IN DIGITAL BREAST TOMOSYNTHESIS
Author:	C. RAVINDRA MURTHY, Dr. K. BALAJI
Abstract:	Digital breast tomosynthesis (DBT) provides considerable benefits over digital mammography for radiologists in terms of cancer diagnosis. In earlier studies, conventional rapid analytical reconstruction techniques in DBT produced noisy images at constrained angles, causing the unconstrained recuperate issue. Consequently, architectural distortion (AD) detection might aid in the early detection of breast cancer, which is currently identified through a manual procedure. As a result, automated identification of AD remains underprivileged. To tackle these issues, a novel Unveil Architectural Distortion with Emendated Image in Digital Breast Tomosynthesis is proposed. To increase the image quality, the input image was first reconstructed using a scaled gradient projection algorithm in a model-based formulation. Following that, the unveil architectural distortion based on slicing process was incorporated to identify architectural distortion in the image. The suspicious spots were detected through a tracks-based trailing approach, and the advanced logistic regression classifier was employed to classify the malignant and benign spots. The results of the implementation show that the proposed approach is effective in detecting breast cancer.
Keywords:	Digital Breast Tomosynthesis (DBT), Architectural Distortion (AD), Slicing Process, Scaled Gradient Projection (SGP), Logistic Regression Classifier, Breast Cancer.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	HOW MARKETERS CAN INCREASE THE RELEVANCE OF EMAIL MARKETING CAMPAIGNS: DATA ANALYSIS WITH MACHINE LEARNING METHODS
Author:	REDOUAN ABAKOUY, EL MOKHTAR EN-NAIMI, ANASS EL HADDADI, LOTFI ELAACHAK
Abstract:	Despite the emergence of many digital marketing strategies, email remains the most effective strategy of all. It generates a high return on investment (ROI) and offers a fast way of retention of customers. Is still the preferred channel, not only for businesses, but also for clients, Convinced of its great capacity for influence, more and more companies are looking for a relevant marketing strategy. Convincing the customer to open the email is the first step to an effective campaign. Therefore, it is important to understand how marketers can improve the open rate of a marketing campaign. Data Driven Marketing encompasses the many techniques and applications that affect digital marketing directly related to the use of customer data. Machine learning can predict user needs based on customer data and past behaviors. Those predictions can then be used to suggest offers that are based on the individual. The objective of the work is to analyze the main factors that increase the opening rate of marketing campaigns. To this end, we are developing classification algorithms to predict whether a campaign will be classified as relevant or irrelevant. To achieve this, we used and evaluated three different classiﬁers. Our results showed that it is possible to predict the performance of a campaign with an accuracy of about 82%, using Adaptive Boosting algorithm and the redundant filter selection technique.
Keywords:	Digital Marketing; Email Marketing; Machine learning, Data Analysis
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	IR-ECLAT: A NEW ALGORITHM FOR INFREQUENT ITEMSET MINING
Author:	MUSTAFA MAN, NUR AQILAH BT RUSLAN, WAN AEZWANI WAN ABU BAKAR, JULAILY AIDA JUSOH, MOHD. KAMIR YUSOF, NUR LAILA NAJWA BT JOSDI
Abstract:	The key challenge of association rule mining (ARM) is to discover and extract a valuable information from databases. Mining valuable information from database could be very challenging especially for decision-making process. This is because mining association rule may require repetitious scanning of large dataset in the databases that can lead to high memory usage and time processing. A few algorithms were introduced by researchers to handle these related problems in data mining. The Incremental Equivalence Class Transformation (I-Eclat), Rare Incremental Equivalence Class Transformation (R-Eclat) algorithm are example of algorithms in rule mining techniques using vertical format data repositories for frequent and infrequent pattern mining. The main operation in I-Eclat and R-Eclat are intersecting tidset. Since the size of tidsets would affect the memory usage and its execution time, more memory and time required for a bigger tidsets. This paper introduces a new incremental of rare pattern mining approach by adopting R-Eclat called Incremental Rare Equivalence Class Transformation (IR-Eclat). The IR-Eclat specifically designed for infrequent pattern mining, and it is beneficial for dynamic database as the data is increasing towards volume with linear proportionate into time. In conjunction with big data explosion, the end users are at an advantage for the use of this incremental approach. The experimental results on several benchmark datasets indicate that IR-Eclat outperforms compare with R-Eclat by reducing its processing time especially in dynamic database as the data is increasing in volume from time to time.
Keywords:	Data Mining, Equivalence Class Transformation (Eclat), I-Eclat Model, R-Eclat, IR-Eclat, Interestingness Measure
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	MACHINE LEARNING-BASED OPTIMAL SEGMENTATION SYSTEM FOR WEB DATA USING GENETIC APPROACH
Author:	N. SILPA, Dr. V. V. R. MAHESWARA RAO
Abstract:	The rapid emergence of computer technology has led to the storage of vast amounts of information in databases. The increasing popularity of electronic data has also created vast amounts of unlabeled information. The potential of extracting valuable knowledge from such digital data has created the basis for the researchers towards important research areas such as Web engineering, Data Science, Big Data Analytics etc. Web engineering is a process utilized for exploring patterns in large databases. However, finding intrinsic structures in large amounts of data becomes a distinctive challenge to organize them into meaningful groups. Many of the existing clustering algorithms are not appropriately suitable for all kinds of web applications. This prompted the present researchers to develop a machine algorithm that is more applicable and robust in real-time to get technological intelligence especially from web data sources. The authors propose an Optimal Segmentation System using a Machine Learning approach (MLOSS) with twin objectives. Initially, MLOSS performs the pre-processing step on the unstructured and semi-structured web documents to prepare efficient data representation structure for applying either supervised and unsupervised techniques. Later as a part of second objective, the proposed system emphasizes on the segment of the preprocessed web data using clustering techniques with an hybridization of the Genetic Approach, that mimic the biological evaluation process having the self-learning proficiency. To validate the performance results of the proposed framework in numerous orders of magnitude, much experimentation is done and, the results have proven this as claimed.
Keywords:	Machine Learning, Web Mining, Genetic Algorithms, Clustering, Un-Supervised Techniques
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	HYBRID RECOMMENDATION SYSTEM TO SOLVE COLD START PROBLEM
Author:	MD MIJANUR RAHMAN, ISMAT ARA SHAMA, MD SIAMUR RAHMAN, MD RAHMATULLAH NABIL
Abstract:	The recommendation system has been very vital in the field of research. The objective of the recommendation system is to recommend items to users, but it is difficult when the user’s purchase history, ratings, personal information are not available. Though many recommendation systems are available to recommend products, it is a big problem for new users because there is no available information that helps to recommend the appropriate products to the new users. To get better enactment of recommendation systems, solving the cold start problem is an important issue for researchers. Many recommendation techniques are available for the last couple of years. It has been overwhelming for the new researchers, merchants, web application developers and etc. to know each of them very quickly. Commonly used possible solutions of coldstart problem, frequently used datasets for the specific domain have not been found. So, various techniques are summarized in this article like hybridization methods, data collection approaches, most commonly used possible solutions of cold start, frequently used datasets, algorithms, evaluation methods etc. This study examines how the cold start problem can be solved by the existing hybrid approaches that may help researchers to get a direction for solving the cold start problem.
Keywords:	Content-Based-Filtering, Collaborative-Filtering, Cold-Start, Hybrid, Recommendation Systems.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	SGD OPTIMIZER TO REDUCE COST VALUE IN DEEP LEARNING FOR CUSTOMER CHURN PREDICTION
Author:	MUHAMMAD DIMAS ADITYA AKBAR, ANDRY CHOWANDA
Abstract:	The number of customers is an important indicator for companies to know the success of a product and service offered. In general, customers are grouped into two categories, loyal customers and disloyal customers. This disloyal customer refers to customers who have stopped using a product and service from a company or are often referred to as churn. For that, we need a system that can predict whether the customer has the potential to experience churn or not. The system is required to predict well with one of the indicators, namely a low-cost value. One way to reduce the cost value is to use the optimizer function. Researchers use deep learning algorithms to create predictive models. A total of 3333 rows and 20 columns of telecommunication customer public data are used as datasets. This study also compares several optimizer algorithms to find the lowest cost value. In addition to the cost value, the accuracy and f1 score results are also used as other considerations. Researchers also use cross-validation for the training process and validating the model created. To deal with the imbalance class dataset, the researcher uses the unweighted method on the cost function. The evaluation results show that SGD has the lowest cost value, which is 0.261. Meanwhile, the AdaGrad matrix classification shows the best value for accuracy and f1 score with 0.926 and 0.846, respectively
Keywords:	Customer Churn Prediciton, SGD Optimizer, AdaGrad Optimizer, Deep Learning, Telecommunications
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	ASSESSING THE USE OF E-PAYMENT IN CULINARY BUSINESSES DRIVEN BY THE TRUST FACTOR IN TECHNOLOGY ACCEPTANCE MODEL
Author:	BELA OKTAVIA, SFENRIANTO
Abstract:	Understanding UMKM's reputation based on merchant acceptance and lack of electronic payments will ensure that your business as a service provider knows the extent of demand. This is the need of small businesses, which can also directly influence the desire to continue electronic payments of UMKMs. Another advantage is that it provides insight and insight to e-payment business players about the service that UMKM wants to achieve based on UMKM's measurable ratings and recommendations, and at the same time Recommendations for quality improvement. This is to present this issue. The study itself aims to analyze the characteristics and factors affecting UMKM when merchants use electronic payment services in West Jakarta using the Technology Acceptance Model (TAM) and is modified by adding an external variable, namely confidence when informing entrepreneurs about their behavior. The results of this study show that the reliability variable has a significant influence on the system.
Keywords:	E-business, E-payment, Fintech, TAM, Trust, UMKM.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	MEASURING RESEARCH INTEREST SIMILARITY AMONG AUTHORS USING COMMUNITY DETECTION
Author:	HARITHA AKKINENI, MYNENI MADHU BALA, VENKATASUNEETHA TAKELLAPATI, MADHURI NALLAMOTHU, SURESH YADLAPATI
Abstract:	Social dynamics that govern human phenomenon are the real need of the hour to access the community structures in social networks. In the present world, online social networks provide huge data that includes the objects information and comments which are analyzed and lead to discovering information and relationship among the networks. Finding community detection is an existing and attracting the researchers where they use different algorithms, one is mathematically based which work on connections in the community and the other one is the graph the structure which shows the output and it is similar to the topological structure. These traditionally followed algorithms and structures are having their limitations. This article attempts to overcome these drawbacks by identifying communities in social networking sites using density-based clustering technique DBSCAN. The identification and removal of such noisy nodes in the identified communities improves the quality. The method's ability to detect different community structures has been demonstrated in studies on synthetic and real-world networks such as research gate, where scientists communicate and share their work and build their reputations.
Keywords:	Density Based Clustering, DBSCAN, Community Detection, Fast Greedy, Visualization
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	EBONN: AN ENHANCED BAYESIAN OPTIMIZED NEURAL NETWORK FOR CLASSIFICATION OF PHISHING ATTACKS
Author:	N.SWAPNA GOUD,DR. ANJALI MATHUR
Abstract:	Phishing attacks became most common cyber security attack in the digital platforms. To identify these types of attacks using URL analysis, the proposed research has identified the important features and then to design a customized neural network, it has enhanced Bayesian optimizer by adding layers and estimator values to find the best parameters that suits for the given dataset. Previous works has focused on implementation using either traditional or pre-trained approaches which has achieved least accuracy with approximate values of 65.25%, so to find the solution for this problem, the proposed research defined an algorithm which defines the objective function to maximize the target values of the estimators. It also implements stratified validation to find the values at iteration and picks the one with best value as output. The model has parameterize all the estimators associated with the neural network and found that it works with an accuracy of 99.5% for training data and with 99.7% for validation data.
Keywords:	Exponential Linear Function, Activation Function, Optimization, Objective Function, Batch Normalization, Pre-Trained Model.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	CROSS-VERSION SOFTWARE FAULT DETECTION MODEL WITH AUTOMATIC DATA SELECTION
Author:	KIRAN JAMMALAMADAKA, NIKHAT PARVEEN
Abstract:	Fault detection in software engineering is gaining attention due to the severe impacts of the faults in the software and malfunctioning of software used in diverse domains. Among the various faults, the class overlapping and the data distribution difference are one of the major issues that are required to be properly addressed. To deal with these issues, we propose a concept on cross-version software fault detection with data selection (CV-FDDS). This scenario is much reasonable and significant to detect the above-mentioned issues by predicting the faults in the new versions based on the labelled data of previous versions. The proposed framework works on identifying the faults in the newer versions of the software projects on the basis ofolder version labels through the use of learning strategies. Initially, the dataset is pre-processed using an appropriate data filter. The older and newer files are treated in two different ways to accomplish this task. To deal with the existing or the older files, an adaptive spiking atom search recurrent neural network classifier (ASARNN) has been proposed through which the training and testing data are automatically selected. This is done by assigning higher weight values to the much relevant and noise-free older versions. Based on the training and testing, the newer versions comprising faults are detected using the proposed kernel optimized extreme learning machine (KOELM). The implementation of the developed approach is implemented on PYTHON. The implemented approach performance is compared against the conventional methods with respect to the major performance indicators. The implemented nodel achievesa geometric-mean (G-mean) of about 0.7815, f-measure of about 0.775 and balance is about 0.8055 respectively.
Keywords:	Fault Detection, Software Engineering, Data Distribution Difference, Class Overlapping, Data Selection, Neural Network.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	MAGNITUDE-BASED TWIN TEXTON CO-OCCURRENCE MATRIX FOR IMAGE RETRIEVAL
Author:	M. VIJAYASHANTHI, DR.V.VENKATA KRISHNA, DR.G.VENKATA RAMI REDDY
Abstract:	The accuracy for any CBIR structure depends on the accurate recovery of relevant images from a larger database. This paper creates a new derivation for existing texton methods to enhance precision in CBIR. No method on textons defined any sub textons primarily based on the magnitude relation between pixels that are part of textons and the rest (not part of the textons). This paper proposed a new variation to texton called magnitude-based twin texton co-occurrence matrix (MTTCM), which explores micro structures of the 2 x 2, and integrates the color, texture, structural features with edge information. The MTTCM is evaluated on natural texture datasets, and the findings show that it outperforms current representative image feature descriptors.
Keywords:	Sub-Textons; Color Features, Texture; Structural; GLCM Features
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	ONLINE CUSTOMER BEHAVIOR IN MOBILE PAYMENT E-WALLET: THE MODEL OF RELATIVE ADVANTAGE
Author:	SATRIO MATIN UTOMO, DONI PURNAMA ALAMSYAH, INDRIANA, LENI SUSANTI
Abstract:	The use of mobile payment e-wallet in Indonesia continues to increase due to the advantages of this payment method. Based on the phenomenon of mobile payment users, this study aims to examine the relationship between smartness, mobility, perceived ease of use, and relative advantage. This research investigates a model that can form the relative advantage of mobile payment e-wallet users. The research method used is a quantitative survey; the survey is carried out on 290 mobile payment e-wallet users in Bandung. The data from users obtain through a questionnaire. Then the data is processed through SmartPLS with several tests, namely the Inner and Outer test and the research hypothesis test. The research variables studied are smartness, mobility, perceived ease of use, and relative advantage, while the relationship between variables is described in the research model. The study results show that smartness has a relationship with perceived ease of use; mobility is said to be able to change perceived ease of use. Perceived ease of use has a positive relationship in increasing relative advantage, while directly smartness and mobility are said to have a positive relationship with relative advantage. Mobile payment e-wallet users seem to give the view that relative advantage is more able to be directly influenced by perceived mobility and perceived ease of use. The perceived ease of use can be well supported through the assumption of smartness and mobility. The findings from the research model explained the ability of perceived ease of use in mediating the relationship between smartness and relative advantage. However, perceived ease of use is not able to mediate the relationship between mobility and relative advantage. The findings from the research are useful for companies in implementing business strategies, where determining business strategies need to understanding user behavior on mobile payments. Furthermore, the research model can be used as an insight in determining marketing strategies to increase the value of mobile payment w-wallet by increasing morning users' relative advantage. The more benefits that the user feels, the more impact the user's interest in using a mobile payment e-wallet.
Keywords:	Smartness, Mobility, Perceived Ease of Use, Relative Advantage.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	HERBAL LEAF RECOGNITION USING MASK-REGION CONVOLUTIONAL NEURAL NETWORK (MASK R-CNN)
Author:	LAIALI ALMAZAYDEH, REYAD ALSALAMEEN, KHALED ELLEITHY
Abstract:	Recent rapid technological advancements in pattern recognition and computer vision have led to great results on a wide range of applications. One of these applications is herbal plant species identification, as the proper automated system for the recognition of herbal plants is required for botanists to study therapeutic and nutritional uses of herbs. In literature, many studies have adopted classical machine learning approaches while some studies have adopted deep learning approaches, via the leaf images. For this work, we use the latest state-of-the-art framework, namely Mask R-CNN, to build such a classification system to identify a medicinal plant. In this paper, we demonstrate the development of the classification system using Mask R-CNN and its backbone: region proposal network, RoI Pooling, RoI Align, and the network head: classification & detection, segmentation. The trained model achieved average accuracy of 95.7% for the identification of 30 medicinal plant species loaded from the Mendely Dataset. The model output is obtained as bounding box for object detection, mask, and class indicating a plant species.
Keywords:	CNN, Deep learning, Mask R-CNN, PPIR, RPN
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	A DETAILED ANALYSIS ON MEDICAL IMAGE PROCESSING TECHNIQUES USED FOR BRAIN TUMOR DETECTION AND CLASSIFICATION
Author:	SAYEEDAKHANUM PATHAN, DR.SAVADAM BALAJI
Abstract:	Detection and classification of tumor portions from the brain images are the challenging and demanding tasks in the field of medical imaging applications. Because, the earlier prediction of tumor is highly essential for the patients to provide proper treatment at the time. So, an automated tumor detection system can be more useful for the medical experts to identify the growth and structure of tumor region. For this purpose, there are different types of medical image processing techniques are developed in the existing works. The main aim of this work is to present the comprehensive survey for analyzing the techniques used for detecting the brain abnormalities. Also, it objects to investigate the operating characteristics, working nature and performance of various image processing techniques. Typically, the preprocessing techniques are mainly used to filter the noisy contents for improving the quality of images with increased contrast. Specifically, the feature extraction models are used for extracting the high level feature attributes and patterns from the filtered image. Then, the optimal numbers of features are selected with the help of optimization techniques by estimating the objective function based on the best fitness value. Here, the importance of using segmentation approaches is to partition the image into collection of pixels, which are helpful for locating the tumor, affected regions. Finally, the classifiers are used for predicting the output label as normal or tumor-affected by training the samples based on the optimal features. For experimental validation, there are different measures have been used to evaluate the performance results of these techniques.
Keywords:	Brain Tumor, Magnetic Resonance Imaging (MRI), Segmentation, Feature Optimization, BRATS Dataset, Deep Learning and Machine Learning Techniques.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	FEATURE SELECTION AND EXTRACTION USING DECOMPOSITION TECHNIQUES IN BIOMETRIC AUTHENTICATION
Author:	Y SURESH, PAVAN KUMAR K, PESN KRISHNA PRASAD, AKKINENI HARITHA
Abstract:	Pattern recognition is a very significant and the most prominent emerging technology which aims at analysis or investigation and construction of pattern. It is highly complex phenomena. Vector logic would provide better strategies and yield results for this problem. Generally, in several phases of human beings life, accurate uniqueness validation seems to be critical. Before the emergence of computing revolution, the issue security is for individual ensured after checking the someone in person and also with the help of the signature. It is observed that conventional methods of authentication are unproductive because any person can impress as a real person with the help of therapeutic procedure and through spoofing. At present, authentication is made through offline and or online mode taking the distinctive features like biometrics of a person. The main use of the distinctive or unique features is that no one can duplicate the features of original human being. In the event of processing the biometric qualities, it is a complex process to derive authentication. In order to enhance accuracy, researchers have proposed diverse types of algorithms. During this process, finger and face traits of a person are considered and also , and applications of Kronecker Product(KP) such as Khatri Rao Product are used. And then, two multimodal authentication systems using AT&T, FERET and Yale data sets are implemented in MATLAB, Python.
Keywords:	Biometric, Khatri Rao Product, Kronecker Product, LU Factorization, Pyparsvd, SVD, LU Factorization.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	THE PERCEIVED EASE OF USE IN MOBILE PAYMENT SUPPORT BY RESPONSIVENESS, SMARTNESS AND MOBILITY
Author:	NORFARIDATUL AKMALIAH OTHMAN, DONI PURNAMA ALAMSYAH, CHYNTIA IKA RATNAPURI, DIAN KURNIANINGRUM
Abstract:	Technology continues to develop, including payment media made online through applications. The presence of technology in payment applications aims to facilitate users and support user performance in their activities. Mobile payment e-wallet is one of the payment media currently increasing in use in Indonesia. Several advantages are considered capable of influencing the use of mobile payment e-wallets. Based on the phenomenon of the problem that occurs, the purpose of this study is to examine factors that can increase the perceived ease of use of mobile payment users, including responsiveness, smartness, and mobility. The research method used in this study is a survey with quantitative data; a survey was conducted to 290 respondents of mobile payment e-wallet users in Bandung (Indonesia). Data from respondents was obtained using a questionnaire; then the data was tabulated and processed by linear regression analysis to find a factor analysis model. The analytical tool used in SPSS was used to process the linear regression analysis also to test the research instrument. The research hypothesis was tested to confirm the factor analysis model. The model tested is then assessed for the level of moderation through gender, namely men and women as one of the characteristics of mobile payment users. Based on the test results, it is known that mobility, responsiveness, and smartness assessed by mobile payment e-wallet users have a positive relationship with the ease-of-use percentage. The essence of this study explains that perceived ease of use can be controlled by supporting factors, namely the mobility of the use of e-wallet, the level of responsiveness of the application, and the view of smartness in using an e-wallet. There is a moderation based on gender, and research findings explain that male gender supports the relationship between responsiveness and mobility, while female gender only supports the relationship between mobility and perceived ease of use. The findings of this study can be used as recommendations for e-wallet service providers in Indonesia, especially banking; several things that are considered by users that need to be considered in mobile payments are mobility, responsiveness, and smartness
Keywords:	Responsiveness, Smartness, Mobility, Perceived Ease of Use.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	UTEM NAVIGATION SYSTEM: PEDESTRIAN AND TRAFFIC SIGN DETECTION USING CNN ALGORITHM
Author:	WAN MOHD YAAKOB WAN BEJURI, MUHAMMAD HARRAZ HARUN, ABDUL KARIM MOHAMAD
Abstract:	Navigation is a common problem for all drivers, especially university visitors. Unfamiliar place making the driver become careless and unaware, which give hazard to pedestrians and driver itself. Thus, this system aims to solve the problems, by developing mobile navigation with safety features by taking UTeM campus as our scope of the study. The system using algorithm using CNN as an algorithm and the architecture used is Tiny-YOLOv2 to detect traffic signs and pedestrians. To begin, the dataset containing Person and Traffic Sign images and their annotations will first need to be acquired. Then, the CNN model will be trained and tested. As a result, our proposed system shows that the mean average precision for both classes can achieve as 90.44%, when it is implemented in a conventional smartphone. This is proof that our system can provide better capability when it is implemented with a smartphone device. Thus, it contributes to being a new mobile navigation system that can provide multiple capabilities, instead of navigation functions. In conclusion, our system was proven to be a valuable solution for the mobile navigation system. In addition, it is implicated to educate the driver community to be a responsible and alert drivers.
Keywords:	CNN Algorithm, Tiny-YOLOv2, Object Detection, Mobile Navigation System
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	ANALYSIS OF FAULT SYSTEM MODEL OF TASK AND RESOURCE SCHEDULER IN CLOUD USING FAULT TOLERANCE SCHEDULING WITH MODIFIED FIREFLY ALGORITHM
Author:	SRIDEEPA.T, DR.BABYDEEPA.V
Abstract:	Cloud computing has evolved into a new user-driven version with easy access to flexible and configurable computational resources such as networks, servers, storage areas, sensible applications, and services, all of which can be accessed concurrently with little need for service provider intervention or control. In general, cloud computing users do not own cloud infrastructure, but rather lease it from third parties to avoid high overhead. The most embarrassing issue in cloud computing is fault tolerance scheduling. An efficient scheduler could improve various aspects of work scheduling in a cloud machine, as well as equal and overall performance. Fault control is used to control the fault tolerance machine without affecting its performance. Even if a count of the machine's losses fails, a fault handling mechanism allows the machine to keep running, albeit at a lower level, rather than failing completely. Various approaches, such as genetic set of rules, ant colony optimization, particle swarm optimization, and so on, have been attempted to solve this problem. The firefly rule set is a metaheuristic optimization rule set that can be easily inspired. Our proposed Fault Tolerance Scheduling, which is based on heuristic algorithms, aims to achieve fault tolerance while also maximizing help utilization in the cloud. The experimental results show that, when compared to existing Dynamic Fault Tolerance Scheduling Techniques (DFTST), and proposed Fault Tolerance Scheduling with Genetic Algorithm (FTSGA), Fault Tolerance Scheduling with Ant Colony Optimization (FTSACO), and Fault Tolerance Scheduling with Modified Firefly Algorithm (FTSMFFA) Algorithms, Fault Tolerance Scheduling with Modified Firefly, and reduces energy consumption.
Keywords:	Cloud Computing, Fault Tolerance Scheduling, Host Active Time, Genetic Algorithm, Ant Colony Optimization, Firefly Algorithm, Energy Consumption
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	A FUZZY LOGIC MPPT BASED CONTROL FOR A PHOTOVOLTAIC SYSTEM
Author:	SALAHEDDINE MOUNTASSIR, SAAD SARIH, ABDELOUAHED TAJER
Abstract:	This paper presents the application of fuzzy logic in the MPPT control of a PV array generation system and a comparative analysis with the widely used P&O algorithm in MPPT in different solar irradiation conditions. These techniques have been improved over time, for example, by improving solar cells, heat transfer liquids, and asynchronous generators, or by improving tracking control algorithms or power electronic components in inverters. The optimization of energy production power from solar photovoltaic systems will be the emphasis of this article. Thus the fuzzy logic control is directly implemented in the MPPT and control a DC-DC boost converter. The simulation is performed using MATLAB Simulink tool and a comparative analysis of results is given with the MPPT based on the P&O algorithm with various steps. The FLC used in this paper demonstrates the relevance of the choice of the input variables and the importance of the use of an optimal number of rules. Those parameters show better performance and faster response than other methods even in changing environmental conditions.
Keywords:	P&O, Fuzzy logic, Boost converter, MPPT, PV system, Matlab / Simulink.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	BIG DATA AND MACHINE LEARNING APPROACH FOR AN EFFICIENT INTELLIGENT LOGISTICS TRANSPORTATION
Author:	Z. MOUAMMINE, H. KHOULIMI, O. EL IMRANI, M. CHRAYAH, A. AMMOUMOU, B. NSIRI
Abstract:	Logistics is the pillar of any industrial activity, transportation and itineraries management are the essential functions of logistics, within the appearance of smart city infrastructure, intelligent transportation appeared as a cutting-edge technological newcomer, though, all researches curried out about intelligent transport/logistics require basically the existence of intelligent ground such smart city infrastructure, sensors or “IoT” so as to work, however, this is still challenging for developing countries. Thus, there is a need for an alternative framework allowing “ITS” to work independently to the infrastructure. To fulfill that demand, authors suggest in this paper an innovative model to enable smarter transportation no matter there is an intelligent ground or not. It enables automatic monitoring of road traffic state and transportation conditions via near-real time detection of road events. Facebook and twitter are used as sources of social big data that are needed for our framework. “PNL” is applied to process Moroccan Dialect texts. The obtained results prove the effectiveness of our approach, to the best of our knowledge; this is the first work treating traffic event detection from tweets written in Moroccan dialect language using machine learning and Apache Spark big data platform.
Keywords:	Big Data, Logistic transportation, Apache Spark, Smart transportation, Machine learning, Support Vector Machine (SVM), Naïve Bayes (NB), Logistic Regression, Social big data
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	USING YOLO IN DETECTING OBJECTS AT MIXFARM OELAMASI
Author:	ALBERT LAMBA , TUGA MAURITSIUS, DRS., MT. PHD
Abstract:	Supervise farm field can be challenging. It takes too much time and effort just to explore the land. So, we suggest the farmer to start using Solar CCTV. Adding AI to detect activity on what happened in live recording monitoring easier. This paper discuss subtopic in artificial intelligence which is Convolutional Neural Network (CNN) to detect object on farm environment using YOLO. This research was carried out in a plantation located in Oelamasi, Kupang Regency, NTT with the aim of identifying object in the farm. This research purpose is to supervise ongoing process on the field. This research in conduct on papaya field with recorded CCTV video. This research used google collab as training platform, Yolov3 as reference model. The available GPU on google collab is Nvidia Tesla K80.We also predesign integrated system using cloud services so the data can be access via mobile apps or web browser. We hope from what we are doing can give others insight about the uses of artificial intelligence then can be applied on industries.
Keywords:	CNN, YOLO, CCTV, CLOUD, Collab
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	ANALYSIS OF INTEREST IN USING E-COMMERCE SHOPEE DURING THE COVID-19 PANDEMIC
Author:	ANNISA NOFITRIANDI, SFENRIANTO
Abstract:	The increasing number of people using the internet encourages a bigger online market. Some marketplaces dominate the Indonesian market, one of which is Shopee. In the current pandemic situation, almost everything the community needs, both primary and secondary needs, can be met from online services. Therefore, there is a demand for shopee companies to know and understand the needs or preferences of consumers. If the company can know and understand the needs or preferences of consumers, the company can win people's hearts to buy on the shopee application. This study aims to find out what factors influence the interest in using e-commerce by customers during the covid 19 pandemic. The method used is SEMPLS analysis, tests carried out by the inner and outer models, which help test the TAM Factor variables, Social Factors, and E-Service Quality. The results obtained Factors that influence attitudes towards the use of e-commerce by customers during the pandemic are factors; Perceived usefulness, Perceived ease of use, Peer Influence Customer Service on attitude toward using in the use of mobile e-commerce applications during the covid 19 pandemic.
Keywords:	E-Commerce, TAM Factor, Social Factors, E-Service Quality, Covid 19
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	FRESH ONLINE TRAINING (FOT) PLAN AND DESIGN MODEL FOR BLACK SOLDIER FLY LEARNING
Author:	KARISMA LINDA NISSA K., SFENRIANTO
Abstract:	Fresh Online Training (FOT) is designed as a platform that educates people in Indonesia. Black Soldier Fly learning is a solution to reduce organic waste during the Covid-19 pandemic in this platform. This study aims to plan and design a model for Black Soldier Fly learning. Based on the research model, there are three steps: analyze customer requirements, plan business model, and design objectives learning. In this study, sample data using a questionnaire to meet customer needs toward online training. Furthermore, Business Model Canvas is a management strategy to visualize business concepts. In addition, ADDIE Model is a guideline in planning, developing, and facilitating online learning. The utilization of the Moodle platform is used to support and build FOT. The platform consists of exciting and valuable features to learn more flexibly and effectively using computers and the internet. By this study, FOT provides new ideas for online training, encourages business strategies in the field of environment, and improves quality to arrange and develop module learning that other researchers have never done.
Keywords:	Fresh Online Training (FOT), Black Soldier Fly, Business Model Canvas, ADDIE Model, Moodle.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	SOCIAL ISOLATION, A NEW VARIABLE AFFECTING BEHAVIORAL INTENTION TO USE SUBSCRIPTION VIDEO ON DEMAND
Author:	FIRMAN YOELIANTO, VIANY UTAMI TJHIN
Abstract:	In this COVID-19 pandemic, there is a change in human behavior. One of the main changes is the increase spending towards entertainment at home such as SVOD services. There is a problem in the industry that is happening in the form of churn, customer is changing from one SVOD services towards the other. There is stiff competition between SVOD services, in order to win this competition company must know what variables affect the intention to use SVOD. This paper aim to find which variables have the most significant effect upon Behavioral Intention to use SVOD using the UTAUT Model and two new variables of Corona Fear, & Social Isolation. This study will use 140 samples of users that have at least try to use Netflix, Disney+ Hotstar, HBO GO, or Viu. The conclusions of this research are that Content and Social Isolation have significant effect upon behavioral intention to use SVOD. This creates several managerial implications for SVOD service companies. First, companies need to increase quantity, quality and suitability of content of the services by either creating new content by working together with other production company or acquire existing contents to be added to the library of contents. Second, companies need to use existing user historical data in order to increase efficiency, this could be done in the form of setting set number of views as a standard, and reviewing it at several interval in order to determine which content should be remove or extended. Third, companies could set a special package for customer when government set social distancing policy, this could be done in the form setting a special price or discount if customer activated their location from set times such as weekends from afternoon until evening, this way customer could still be engaged towards the SVOD services even in isolation.
Keywords:	SVOD, UTAUT, COVID-19, Streaming, Lockdown
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	IMPLEMENTATION OF DATA MINING TO DETERMINE STUDENT MAJORS USING THE MACHINE LEARNING
Author:	BERNADETE DETA, TUGA MAURITSIUS
Abstract:	Implementation of evaluation, planning, and decision making can be done better if an organization has complete, fast, precise, and accurate information. The required information can be retrieved from operational data stored in the integrated database. Data mining has an important role in everyday life. By understanding its definitions, functions, methods and applications, it will be easier to put them into practice. In addition, data collection is also very much needed in various fields of life ranging from telecommunications, insurance, sports, finance, academic fields and other fields. In-depth understanding of data mining is needed to simplify the work. This study examines the extraction of operational data and then analyzes the data using data mining techniques. Data Mining is the process of analyzing data using software. To find certain patterns or rules from a large amount of data that is expected to find knowledge to support decisions. This study uses student data which includes data on National Examination scores, written test scores, interest and aptitude test scores to determine student majors. In this study, the data mining technique used is Classification. The way to do the classification is by using data mining techniques using machine learning. In this study, the data mining technique used is classification using the CRISP-DM method and modeling by comparing the four models namely Decision Tree, Naïve Bayes, KNN Classification and Random Forest with Rapidminer tools to help find characteristics or variables that support in determining student abilities. . Furthermore, it can be used for future student majors. From the results of the analysis that has been done, the model using Decision tree has an accuracy of 95.58%, Naïve Bayes 88.97%, KNN Classification 93.54% and Random forest has an accuracy of 96.46%. The final conclusion is that modeling with Random Forest can be used to help determine student majors in the best private catholic high school on the island of Flores.
Keywords:	Classification, Data Mining, CRISP-DM, Student Majors, Decision Tree Algorithm, Naïve Bayes, KNN Classification, Random Forest, Mechine.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	BIG DATA FOOD MAPPING FRAMEWORK USING MAP REDUCE TECHNOLOGIES FOR EFFICIENT ONLINE FOOD MARKETING SYSTEM
Author:	PRAJNA HEGDE, DR. PRASHANT BHAT
Abstract:	Companies with a strong customer emphasis are the primary users of data mining nowadays. Inconsistencies and redundancies are not immediately applicable for starting a data mining process since data is likely to be faulty. In data analysis, missing values are unavoidable, and cause serious problems. Faulty knowledge is retrieved and incorrect inferences are drawn when missing values are handled and replaced incorrectly, thereby leading to high computational time with high data and large iterations. Due to heterogeneous datasets, there is a lack of exact mapping technique, and while visualizing the extracted data the mild handling of data leads to inaccurate data extraction, which lowers all the efficient processes contributing towards the accuracy. Hence in this paper, to overcome the issues a novel Big Data Food Mapping framework is proposed which incorporates association rule mining, followed by a fuzzy clustering-based binning procedure followed by a score-based normalization. Map Reducing, including initial mapping based on genetic operators and entropy generation, was also performed. The reduction is done using fuzzy logic and a Graph Neural Network. After that, the data is mined using Concept analysis and Long Short –Term Memory (LSTM). The outcomes were more accurate and displayed using T-distributed stochastic neighbor embedding (t-SNE) based dual clustering.
Keywords:	Online Food Delivery, Fuzzy Clustering-based Binning Technique, Score–based Normalization, Map Reducing, T-distributed stochastic neighbor embedding (t-SNE), Dual Clustering.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text

Title:	IDENTIFICATION OF BANKNOTES ON THE VISUALLY IMPAIRED PERSON THROUGH DEEP LEARNING
Author:	EDDY MUNTINA DHARMA, AGUNG TRISETYARSO
Abstract:	The identification of banknotes is one of the main problems for the blind person. Money is a tool that is used in everyday life to carryout buying and selling transactions by all humans in every part of the world, so this makes money as a primary item for everyone, even for people with disabilities such as the visually impaired ones. The weakness of the visually impaired person in seeing and identifying money can cause into money confusion, mis-taken, or even deceived at the time of the transaction. Therefore, tools are needed to ease the visually impaired to identify the value of money. The purpose of this study is to propose one of the deep learning-based methods of convolutional neural networks (CNNs) that can be used to detect the value of Indonesian currency banknotes. Data sets used in the form of paper money images with an amount of 1000, 2000, 5000, 10000, 20000, 50000, 100000 thousand rupiah, which amounted to 100 images each, bringing the total images used as a dataset reaching 700 images. At the training stage, the CNN model that was built received 560 paper money image inputs (80% of the total dataset), while, 140 images used at the testing stage (20% of the total dataset). Once the trial completely conducted, it is obtained 94. 29% as the best accuracy in the 60th epoch with the kernel size in the first convolutional layer is 3x3 and the kernel size in the second convolutional layer is 2x2.
Keywords:	Deep Learning, Convolutional Neural Networks (CNN), Banknote, Visually Impaired Person.
Source:	Journal of Theoretical and Applied Information Technology 15^th June 2022 -- Vol. 100. No. 11 -- 2022
Full Text