		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

April 2023 | Vol. 101 No.7

Title:	OPTIMIZED CONTROLLER BASED VOLTAGE QUALITY ASSESSMENT IN GRID CONNECTED HYBRID MICROGRID
Author:	M.DEVIKA RANI, V.SAI GEETHA LAKSHMI, P.MUTHU KUMAR, D. R. BINU BEN JOSE, R. SARAVANAN
Abstract:	Growing distributed generation based on renewable energy sources pushed for the improvement of power quality. The variable nature of interconnected renewable energy sources such as wind and solar generators, which are reliant on weather and climate change, has an impact on the power quality of micro grids. Power quality assessment involves a number of metrics; including voltage quality, voltage imbalance, sag score, and current score (current THD). Good power quality assessment reduces the energy losses in a power system resulting in high profitability. In this research; voltage quality assessment in hybrid microgrid using a novel krill herd optimization (NKHO) technique is described. Use of NKHO in voltage quality assessment for hybrid microgrids offers a powerful and efficient means of optimizing the control and operation of the microgrid, ensuring its reliability and minimizing its impact on the grid. The proposed technique enables the identification of sensitive buses and the optimal sizing of energy storage systems to mitigate the impact of voltage sags. This research assessed the application of a fractional order proportional, integral and derivative (FOPID) controller for microgrid voltage regulation. The proposed hybrid microgrid is developed using Matlab/Simulink environment.
Keywords:	Distributed Generation, Power Quality, Microgrid, Novel Krill Herd Optimization NKHO, Fractional order PID Controller FOPID
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	TWO PHASE WEIGHTED ADAPTIVE MEAN FILTER FOR HIGH DENSITY IMPULSE NOISE REDUCTION
Author:	MANOHAR ANNAPPA KOLI
Abstract:	This Paper presents highly efficient high density impulse noise removing algorithm by combining good the features of impulse noise removing algorithms. This paper presents two step noise reduction process. In first stage efficient noise detection algorithm is used to detect the noisy pixels present in corrupted image. In second stage features such as distance and directions of pixels are used in calculating the replacement values of noisy pixels. Efficiency of algorithm is tested by comparing with efficient existing algorithms. An experimental result shows that the proposed algorithm works with very high density noise up to 98% of noise ratio.
Keywords:	Impulse Noise, Adaptive Median Filter, Image Enhancement, Noise Detection.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	THREE TIER FRAMEWORK IRIS AUTHENTICATION FOR SECURE IMAGE STORAGE AND COMMUNICATION
Author:	P. ALLI, J. DINESH PETER
Abstract:	Due to the increasing popularity of multimedia technology, the need for secure image storage and communication has become more critical, because strangers try to access these data for illegal uses. Most researchers try to provide secure image communication and security frameworks, however, they have some limitations like high cost, minimum level of security, authentication issues, etc. To overcome that issues, our proposed work provides a three-tier architecture with secure iris authentication, secure data storage, and communication. Tier 1 includes user authentication with multifactor, authentication factors are username, password, mobile number OTP and iris authentication. Tier 2 has image encryption technology through a two-fold map concept known as the Twofold Logistic Chaotic Map, here the keys are generated using a pseudo-random number (PSNR) generator for randomness. Tier 3 has the communication phase, if two parties need to communicate between them, quantum key distribution along with PSNR is implemented to ensure secure communication. Finally, the proposed method was subjected to various experiments for performance analysis, including a histogram, an entropy rate, a number of pixels change ratio, and a correlation coefficient, through the analysis of the key space, the method can improve the security and reliability.
Keywords:	Quantum Key Distribution, Image Communication, Logistic Chaotic Map, and Pseudo-Random Number
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	LWC: EFFICIENT LIGHTWEIGHT BLOCK CIPHERS FOR PROVIDING SECURITY TO CONSTRAINED DEVICES A SOLUTION FOR IOT DEVICES
Author:	ASHU ABDUL, GARLAPATI NARAYANA, R. SUDHA KISHORE, B. SRIKANTH, K. KRANTHI KUMAR, D.N.V.S.L.S. INDIRA
Abstract:	Internet of things (IoT) is the infrastructure of global network for the information to the nation for societal use and enabling smart services by interconnecting virtual and physical devices or things based on previous, current and future technologies. IoT application is important to people yet in case the IoT system can't safeguard the customer data from software engineer, attacks, and shortcomings. Lightweight encryption is a space of a customary cryptographic estimation that is proper for resource obliged contraptions in IoT. Related work for lightweight techniques used for secure data transmission is portrayed in this paper. Security in IoT is still difficult task, to address security issues Lightweight Cryptography Techniques were introduced and to answer security aspects here the paper is going to present some techniques PRESENT, and its equivalent Methods. The term Lightweight came into picture when Lightweight wireless technology is used to run Lightweight IoT devices. Because the sensors used in IoT Devices are low power and less weight so the need of low power, less weight became reason to create Lightweight wireless technologies. This paper discusses major security challenges of IoT devices besides the performance evaluation of various Lightweight cryptographic algorithms.
Keywords:	Lightweight, Cryptography, Security, Iot Devices, Block Ciphers, Lightweight Protocols.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	CYBER SECURITY THREATS DETECTION AND PROTECTION USING MACHINE LEARNING TECHNIQUES IN IOT
Author:	MS. PRAGATI RANA, DR. B P PATIL
Abstract:	Recently, technology has enhanced itself to the 4th Industrial Revolution, with the Internet of Things (IoT), Edge computing, Computer safety, and along with Cyber-attacks are rapidly evolving. The quick proliferation of Internet of Things (IoT) devices and web in many shapes produces more data, posing cyber security risks. Detection and protection of cybersecurity threats is a significant concern in IoT. Machine Learning (ML) methods are widely regarded as one of the most promising solutions to address cyber security threats and provide security. Machine Learning (ML) methods are crucial in various cyber security applications. This study examines the literature on Cyber security threat detection and protection in IoT such as detection of spam, malware and intrusion over the previous ten years using machine-learning methods. The scope of Systematic Literature Review includes an in depth examination of the majority of ML trending methods in cyber security threat detection and protection in IoT. In recent years, increased Machine Learning techniques are used to solve four major cyber security issues namely identification of Intrusion, Android malware, Spam and Malware.
Keywords:	Cyber security, Threat Detection, Security Risks, Machine Learning, IOT
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	CREATING THE BEST DIRECTED RANDOM TESTING METHOD TO MINIMIZE INTERACTIVE FAULTS- EMPIRICAL PERSPECTIVE
Author:	T BALAJI, P.RAVI KUMAR, M.V.GANESWARA RAO, GEETHA DEVI APPARI
Abstract:	Here we are mainly concerning the problem of randomly generated test cases. Randomly generated test cases will contain some ambiguous test cases, which leads problem at organizational level. A random algorithm will generate random test cases each time, which will contain some similarity on each time. Another problem related to random algorithm was of time consuming process. To removing these issues we proposed our new technique, which will reduce the given drawbacks. We proposed an Adaptive Genetic Algorithm (AGA) which will provide legal input in each case it applied. Thus the problem of ambiguity will decrease. In this research, the optimal inputs will be generated based on Adaptive Genetic Algorithm (AGA) which will reduce the illegal inputs and equivalent inputs. The fault detection rates will be the fitness of AGA. To reduce the fault proneness, AGA uses the coverage metrics of the test cases
Keywords:	Random Testing, Aga, Metrics, Interactive Faults, Empirical Study
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	CHARGING STATION FOR ELECTRIC VEHICLE USING HYBRID SOURCES
Author:	B. MOHAN, M.V.RAMESH, P.MUTHU KUMAR, RAJAN. VR
Abstract:	In future transportation depends on Electric Vehicles (EVs). EVs are run by rechargeable batteries which consume more amount of time to get recharged. Based on voltage level and current rating, EVs have 3 levels of DC fast charging. Each level of charging system has various limitations and takes more time. EVs are not preferred by the consumers as their charging time lasts longer and range anxiety. So, to overcome the issues a combination of the solar and grid supply is used to get fast charging to batteries. The charger uses a DC to DC converter which enhances the current and voltage ratings, so the charging time of the battery will decrease. The interconnected solar powered and grid connected hybrid source charging system was developed for EVs charging. The characteristics of Lead-acid and Lithium-ion batteries were studied. Obtained results shows that Lithium-ion battery have better performance. The results are verified by using MATLAB/Simulink Software.
Keywords:	Solar Energy, EV Charging station (EVCS), Maximum Power Point Tracking (MPPT), DC-DC Converters, Permanent Magnet Brushless DC (PM BLDC) motor
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	E-WALLET ADOPTION IN INDONESIA: REALIZE E-WALLET GROWTH PROJECTIONS UNTIL 2025
Author:	GIOVANI ANGGASTA PURNAMA, NILO LEGOWO
Abstract:	Until now, the use of e-wallets in Indonesia has continued to increase year after year, so it is not surprising that there are projections stating that its growth will reach 3 times until 2025. Of course, e-wallet service providers must improve their quality. Thus, this research aims to determine what factors influence a person's adoption of e-wallets to make digital payments in Indonesia using the updated DeLone & McLean IS Success Model, which has been modified. It gathers data from 401 respondents. The collected data were analyzed using Partial Least Square Structural Equation Modeling (PLS-SEM) on SmartPLS 4 software. The results found that InQ, SyQ, and SeQ have a significant effect on PT; InQ, SyQ, and SeQ have a significant effect on Sat; PS, InQ, SyQ, and SeQ have a significant effect on ITU; PT and Sat have a significant effect on AoE. Apart from the variables above, it has no significant effect on other variables. In connection with the research results, e-wallet service providers must improve the quality of information, systems, and services to gain the trust and satisfaction of their users because this will affect the adoption of e-wallet use. In addition to improving the quality of information, systems, and services, it is also necessary to increase security guarantees that can increase the intention to use e-wallets. The findings of this research can serve as a guide for e-wallet service providers to build a better e-wallet development plan.
Keywords:	Digital Payment, Adoption of E-Wallet, Updated IS Success Model, PLS-SEM, Indonesia
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	ADAPTIVE SCALING IN DEEP CONVOLUTIONAL NEURAL NETWORKS FOR COVID-19 IDENTIFICATION IN X-RAY IMAGES
Author:	MOHAMMED SIDHEEQUE, P. SUMATHY, ABDUL GAFUR. M
Abstract:	For X-ray image COVID-19 identification, researchers turn to deep convolutional neural networks (CNNs). Deep convolutional neural networks (CNNs) may struggle with detecting objects in noisy X-ray pictures. To enhance COVID-19 identification in noisy X-ray pictures without the need for any pre-processing for noise reduction, we offer a unique CNN technique using adaptive convolution to increase the resilience of a deep CNN against impulsive noise. Based on the standard CNN architecture, this method adds an impulsive an adaptive convolution layer, adaptive scaling layer, and a noise-map layer. Furthermore, we enhanced the deep CNN's generalization by using a learning-to-augment technique on noisy X-ray pictures. Twenty-nine hundred and three chest X-ray photos have been gathered, including those showing Healthy and COVID 19. GoogleNet, SqueezeNet, ResNet18, MobileNetv2, ShuffleNet, ResNet50 and EfficientNetb0 are just some of the pre-trained networks that have had their design tweaked to improve their resilience against impulsive noise. Verification using a distorted X-ray ResNet50 trained with the suggested noise-robust layers and learning-to-augment technique outperformed the state-of-the-art method by 2%.
Keywords:	Covid-19 Classification, Adaptive Resize, Data Augmentation, Noise, Adaptive Convolution
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	TECHNOLOGY FOR OBTAINING INFORMATION ON CO2 EMISSIONS CAUSED BY NATURAL AND MAN-MADE FACTORS USING THE ARIMA MODEL
Author:	MAYYA DUBOVIK, SERGEY DMITRIEV, ARTEM M. SHAPOSHNIKOV, GULCHEHRA XALMATJANOVA
Abstract:	The study aims to test a hypothesis on the applicability of the ARIMA model for forecasting time series that reflect the dynamics of carbon dioxide emissions into the atmosphere caused by natural (the Mauna Loa volcano) and anthropogenic (CO2 emissions in Germany, France, and Italy) factors. Stationarity of the time series is tested using the augmented Dickey-Fuller test; autocorrelation of the time series is tested using the Ljung-Box test. The study identifies forecasted values of CO2 emissions for stationary and non-stationary time series. In the first case, the obtained forecasted values of the time series are more precise. The authors conclude that additional adjustments are needed to increase the predictive capabilities of ARIMA models for non-stationary time series.
Keywords:	CO2, Climate Change, ARIMA, Stationary and Non-Stationary Time Series, Forecasting
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	DESIGN AND BUILD A MICRO INFLUENCER RECOMMENDATION SYSTEM AS A SOCIAL MEDIA PROMOTION TOOL USING A SIMPLE ADDITIVE WEIGHTING ALGORITHM
Author:	ADINDA RAMADHANI WIBOWO, ANGGA ADITYA PERMANA, YAMAN KHAERUZZAMAN
Abstract:	Social media usage globally has change marketing strategy. It has become an important marketing strategy to build business. One of the social media is promotion through influencers on Instagram. Customers probably made decision as a result of watching influencers on social media. Business owners who promote their business using social media often feel confused about choosing the right micro influencer to promote their business. Micro influencer is the level of influencer who has followers ranging from 1,000 to 100,000. However, choosing the right micro influencers is not easy task because more people gaining number of followers in social media. Finding the right micro influencer becomes crucial, therefore we proposed a website to choose the right micro influencer according to their preferences. Decision support system to facilitate micro influencer decision making process was named a micro-influencer recommendation system using the Simple Additive Weighting algorithm. This algorithm Simple Additive Weighting was chosen to determine the weight value of each criterion that will select the best alternative from a number of alternatives and the assessment will be more precise because it is based on predetermined criteria values and preference weights. Testing has been carried out to ensure that the algorithm used runs as it should. User testing has also been carried out and has a final satisfaction value of 87.41 % using the End User Computing Satisfaction method.
Keywords:	End User Computing Satisfaction, Micro-Influencer, Simple Additive Weighting, Recommendation System Website
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	SOCIAL MEDIA APPLICATIONS FOR MSMES IN THE ERA OF THE DIGITAL ECONOMY
Author:	MEIRYANI, SUHONO HARSO SUPANGKAT, NI PUTU AYU VANIA VIDYAWATI, SHI MING HUANG, GATOT SOEPRIYANTO, DIANKA WAHYUNINGTIAS, AGUSTINUS WINOTO
Abstract:	Social media provides opportunities for new ways of doing business by sharing, collaborating and reinventing. However, information on how MSMEs utilize social media in MSME trading is still very limited. Therefore, the purpose of this study is to provide a description of the role and benefits of social media applications in SMEs trading in the digital economy era. In a progressively advanced time, the improvement of innovation has made major changes in the business world. One of them is on social media which is progressively creating and facilitating communication. Communication in trade is exceptionally critical. It is essential to realize the existing conveniences and got to be utilized as well as conceivable. This study sought to examine of development of social media applications for MSMEs in the digital economy era: A Case of SMEs at South Jakarta City – Ulujami. By utilizing the meet strategy to thoroughly discuss the utilize of social media in their trade, and utilizing supporting questions, the comes about of the study show that the utilize of social media can be made to encourage media advancement, so as to make strides the development of MSMEs themselves.
Keywords:	Social media, MSMEs; Promotion, Innovation
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	THE ASSESSMENT OF USABILITY OF SOCIAL MEDIA SITES: A HEURISTIC EVALUATION
Author:	MAHESWARA RABBANI, DIYURMAN GEA
Abstract:	One of the success factors of a website is the quality/level of usability. In user-centered design, a high level of usability is required to create harmony when users interact with the product. The purpose of this research is to evaluate and develop the potential usability interface of a social media site using the heuristic evaluation method. The study's results found that there were 33 problems and 33 recommendations for user interface improvements. Other results show that the most significant usability problem lies in the 'User Control and Freedom' area, with the lowest severity rating, 3.20, compared to the overall average severity rating, which is 3.47. On the other hand, the 'Match Between System and the Real World' variable gets the highest severity rating with a value of 3.58. The findings in this study are applied to improve the usability and user experience of a website and as a reference for supporting effective interface designs in the future.
Keywords:	Heuristic evaluation, Severity rating, Usability, User experience, User interface
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	STATE OF ART DIGITAL LEARNING: PROPOSED FRAMEWORK, METHODS, BENIFITS, TOOLS, AND CHALLENGES
Author:	SUNEET WALIA, RANJIT RAJAK, SATISH CHOURASIYA, OM PRAKASH KAIWARTYA
Abstract:	Covid-19 has enabled a paradigm shift in the teaching-learning process. It has instilled the need for a platform that does not bound any learner to a particular campus or four walls of a classroom. Education for the 21st century endures encouraging discoveries in the area of digital teaching-learning. Moreover learning in present times has engendered different platforms for students and teachers. The most prominent one in the present era is the virtual platform for interaction and learning. But digital learning platform is marked by problems of massive expenditure on computer resources such as servers, network devices, etc. which further require maintenance and technical manpower for breakdown resolution. This is made easier through a cloud computing environment. This paper endeavors to explore the various digital learning methods, potential benefits, and core challenges of digital learning. Further, an endeavor is made to propose of framework describing the usages and key benefits of cloud computing for teaching-learning purposes
Keywords:	Digital Learning, Digital Environment, Tools, LMS, MOOC, Cloud Computing)
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	AN EVALUATION OF A RESOURCE ALLOCATION AND SCHEDULING MODEL SMART GRID ENVIRONMENT IN DIFFERENT LOAD AND FAULT SITUATIONS
Author:	NISHANT JAKHAR , RAINU NANDAL , RITU NANDAL
Abstract:	The power clamor of consumers is escalating steadily; Resulting surge on electric depots that leads to subsequent electric collapse. One of the produced distresses of power zone is the uneven and rising power demands. An equilibrium of power dispensation with respect to consumer’s requirement is trailed by smart grid in traditional dispersal process. Despite the fact that bulk load conditions amid constant power reserve; are still an obstacle in smart grid system. The presented study imparts stipulated and load governed technique to get the competent source allotment &utility. The suggested smart system will anticipate the probable strain during allotment time and will function load evenness utilization of accessible assets. The said framework is evaluated in this research under three different load conditions called Low, Average and High. The obtained outcomes are for mean energy hinderance, power controls and power collapse estimations. The introspected consequences determined the suggested application based decisive paradigm that enhanced the conduction of Intelligence system during utmost burdened times & attained productive solution within shorter span of time and minimum power collapse. The brought-forwarded technique gained genuine and efficacious transmission of the energy/power.
Keywords:	Electric Collapse, Power Zone, Source Allotment, Application Based, Intelligence System
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	USING DATA MINING TO INFLUENCE SOCIAL ENTREPRENEURSHIP AND TERRITORIAL DYNAMICS
Author:	JIHANE MAFTAH, SALMANE BOUREKKADI, YOUSSEF PHENIQI
Abstract:	Economic digitization is a trend phenomenon that shifts microeconomics and macroeconomics, in other words, the economy from the era of the physical world or material space to a virtual world or immaterial area by relying on digitization technologies, in volumes never before achieved and in record time, directly impact the performance of organizations as well as social entrepreneurship in various territories around the world. Organizations must therefore deal with a gigantic wave of data to follow the news as well as promote territorial intelligence. This universe of data is defined by the generic term of Data mining, the processing of which is usually provided by artificial intelligence technologies. Entrepreneurship constitutes a space of solidarity, mutual aid, pooling of social groups who suffer from poverty, precariousness or even and aspire to values such as dignity, interests and power and observed in order to ensure an equitable distribution for the benefit of all citizens and regions of the Kingdom. It is within this framework that Morocco has embarked on a structural transformation which has resulted in the efforts of the State to strengthen a modern and competitive economy, to simplify the creation of their growth and the promotion of entrepreneurship and by therefore promote the development of the territory. In the interest of having a much broader and clearer vision of the influence of data mining and business intelligence on territorial intelligence and social entrepreneurship, We are trying to answer the following problem: How does social entrepreneurship promote a territorial dynamic using Data mining and Business intelligence, particularly in the Fez-Meknes region?
Keywords:	Data Mining, Business Intelligence, Social Entrepreneurship, Smart Territories
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	BOOST STOCK FORECASTING ACCURACY USING THE MODIFIED FIREFLY ALGORITHM AND MULTICHANNEL CONVOLUTIONAL NEURAL NETWORK
Author:	NILESH B. KORADE , DR. MOHD. ZUBER
Abstract:	The number of stock investors is growing every day, so providing more accurate predictions for trend and stock value is essential to earning money. Neural networks have been used in different research studies to predict future movements in stock prices. In comparison to other methodologies, convolutional neural networks (CNN) perform impressively in stock price forecasting. A parameter whose value is used to regulate the learning process is known as a hyperparameter. One of the main issues in CNN is selecting the set of appropriate parameters to train it on a stock dataset that gives correct results. Due to the stochastic nature of a neural network, which uses randomness during learning, there are variations in the performance of CNN on the same dataset. In this study, we proposed a modified firefly algorithm (MFA) approach to identify the best configurations for CNN architectures, and a multi-channel convolutional neural network (MCNN) is used to improve CNN accuracy in stock forecasting. The MFA controls the random movement of fireflies by updating the randomization parameter, and if there is no local best solution in the neighborhood, the firefly will move towards the best solution instead of moving randomly. The outcome of the comparison between the proposed method and existing methods demonstrates that it has the potential to identify the best CNN design parameters in fewer iterations and forecast stock value more accurately, which is closer to actual value. The MFA-MCNN performs better than several state-of-the-art optimization techniques and offers higher accuracy. A smaller number of iterations is needed by the MFA-MCNN to determine the ideal parameter for the CNN architecture. The randomization parameter has a high value in the early iterations, forcing MFA to find the best solution; but, as the iteration progresses, the randomization parameter's value decreases, providing greater convergence.
Keywords:	Convolutional Neural Networks, Forecasting, Firefly Algorithm, Optimization, Stock.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	SOFTWARE/HARDWARE TASK MIGRATION BASED ON A CHECKPOINT FOR MULTIPROCESSORS EMBEDDED SYSTEMS
Author:	KAMEL SMIRI, FAYCEL EL AYEB
Abstract:	Traditionally, the decision of implemented particular task in hardware or as software program running is typically taken at design time. This partitioning is static which has huge effect to the performance of a system. In the related work on Hardware/Software Co-design, the Hw/Sw partitioning problem is solved offline by mapping each task to dedicated core (Hard or Soft) with respect to certain objectives such that throughput, power consumption. But at now days and with the introduction of reconfiguration at runtime, the Software/Hardware partitioning problem can be solved online, thus, taking advantage of Hardware high performance and Software flexibility with the lowest possible costs. In this context, our approach focuses on these objectives. However, for a wide range of embedded application, the cohabiting software flexibility and hardware speed is a key feature to provide performing and flexible embedded systems. While in this system exit some tasks that are appropriate for a software program running on a processor and other tasks that might have highly throughput requirements that can be executed by dedicated hardware modules.
Keywords:	Dynamic Migration Of Soft/Hard Tasks; Embedded Systems; Checkpoint; Soclib; Gaut.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	A TRUSTED NODE FEEDBACK BASED CLUSTERING MODEL FOR DETECTION OF MALICIOUS NODES IN THE NETWORK
Author:	SIMHADRI MADHURI, DR. S VENKATA LAKSHMI
Abstract:	The study of malicious nodes is an interesting one, but it has had a negative impact on network performance characteristics. A degraded network response time may be caused by the attacker node's impact on network throughput. Analytical methods have been employed to combat the problem. The type of node is determined by analyzing the existing node's activity and its properties using behaviour based detection for malicious node detection. In the guise of machine learning, intelligent systems are devising new methods for locating and eradicating malicious nodes from the system. Lowering the severity of the data transmission degradation will be a difficult task because malicious nodes share many of the same properties as trusted nodes in the fixed region. Due to an increase in the number of malicious nodes, network performance will suffer. Malicious nodes in the network can affect metrics such as packet delivery ratio, performance, detection rate, energy consumption, accuracy value, and link failure. The proposed model calculates the trust factor of every registered node in the network. The trust factor is used in the process of node authentication and in detection of malicious nodes in the network. The proposed primary security module includes a dynamic authentication mechanism that allows current nodes to authenticate incoming new nodes, resulting in the development of secure links and disseminate authentication between surrounding nodes. The authentication strategy prohibits external hostile nodes from gaining access to the system. A Trusted Node Feedback based Clustering model for Malicious Node Detection (TNFC-MND) is proposed in this research for the detection and removal of malicious nodes in the network. Each node in the cluster receives the cluster key from the cluster head, and this key is used to exchange data with the cluster head. Every time a node sends data to the cluster, the cluster head verifies this key to see if it matches the cluster table. It will only acknowledge this node as a member of the cluster when the match is valid; otherwise, it will be deemed malicious. The proposed model is compared with the existing model, and the results exhibit better performance.
Keywords:	Network Nodes, Node Behaviour, Trusted Node, Network Cluster, Cluster Head, Malicious Node, Node Feedback, Data Loss
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	FACTORS THAT INCREASE INTEREST IN USING MUTUAL FUND APPLICATION WITH USER ACCEPTANCE APPROACHES
Author:	RAYMOND KANDIFAR, ASTARI RETNOWARDHANI
Abstract:	There are many online investment applications that are not registered to OJK (Financial Service Authority) so that the legality of the application does not exist. This legality makes the credibility of other online investment applications highlighted. Credibility affects the intention to use an application. A mobile application success is also depending to intention to use. These issues should be highlighted because give direct impact both for legalized mutual fund application and the citizen. The purpose of this study is to determine the factors that affecting the intention to use mutual fund investment applications Bibit.id in the millennial generation. This Study use Technology Acceptance Model (TAM) that has modified by the Author. The modifications of Technology Acceptance Model using several external variables such as Financial Literacy, Trust, Social Influence, User Interface, User Experience. The results of the analyzed data show Financial Literacy, Perceived Ease of Use, Social Influence, User Experience, and Perceived Usefulness should be a consideration while develop the mutual fund application. In this research found Financial Literacy have positive impact to Intention to Use, more easily the application used makes the application more useful. The useful application affecting to intention to use. Social influence both working environment or family makes the application more easy to use. Users feel, the easier to use application better than the application interface.
Keywords:	Intention to Use, Mutual Fund Application, Bibit.id, Technology Acceptance Model
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	ENHANCED PREDICTION OF CROP YIELD USING DYNAMIC ANT COLONY OPTIMIZATION-BASED RECURRENT NEURAL NETWORK (DAC0-RNN)
Author:	SHANMUGA PRIYA S ,DR.M.SENGALIAPPAN
Abstract:	Machine learning (ML) based crop yield prediction is a method of using classification algorithms to analyze and predict the yield of crops based on historical data and current environmental conditions. Crop yield prediction models require a large amount of data for training and validation, such as weather data, soil data, and historical yield data. ML-based crop yield prediction aims to improve crop yields by providing farmers with accurate and timely information about their crops. This paper presents a new deep learning strategy called Dynamic Ant Colony Optimization-based Recurrent Neural Network (DACO-RNN) to address the issues of low classification accuracy often encountered with traditional ML algorithms during training and testing phases. The DACO-RNN is inspired by the foraging behavior of ants in nature and utilizes an optimization technique to improve the performance of the RNN, which otherwise would degrade during the back-propagation phase. The effectiveness of the DACO-RNN was evaluated using two crop yield datasets and standard ML metrics, and the results showed that it had better performance than existing classifiers.
Keywords:	Crop Yield, Prediction, Classification, Ant Colony, Neural Network, Optimization
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	IOT-BASED COMPUTATIONAL INTELLIGENCE TOOL FOR PATERNITY TESTING AND PARENTAL COMPARISON USING CORRELATION-BASED CLUSTERING AND SVM CLASSIFICATION
Author:	DR.VIJAY ARPUTHARAJ J, DR. K. SANKAR, DR. KUNCHAM SREENIVASA RAO, DR. G.N.R. PRASAD, MR. R.BHARATH KUMAR
Abstract:	DNA Paternity testing and parental comparison research is a method of stifling DNA grouping to pinpoint techniques for identifying the traits of character, setup, nature, and attributes. In this modern era, computational intelligence has proven to be a vital tool that interprets big biological data. It has big impact in the fields of molecular biology and DNA sequencing applications. This technology of IoT is able to determine best results among the big data in concise time with no errors which contribute to the bioinformatics field and researchers. By using correlation-based clustering and Modified Naive Bayesian Classification to analyze quality succession information, it is possible to separate the detrimental characteristics of diabetes from the vast array of DNA quality arrangement components that are included in the collection of copious quantifiable data. This process aims to validate, choose methods and tools for examining poor quality successions. Additionally, it aids in the accurate and serious characterization and translation of outcomes. For information assessment, this study combines regulated and solo IoT based methods. Although the order is completed by MNBC processes, CBC completes the grouping. Health disorders and their physiognomies are associated with a person's gene expressions in genomics and medical sciences. This analysis of correlation based and SVM classification has a massive influences and applications in genomic sequencing and gene mining. The objective of this research is to identify various gene sequences in biomedical inherent learning especially for data related to paternity analysis and testing. The domain and sub domain used here are analysis of correlation based clustering and SVM classification for gene sequence data analysis respectively. The proposed technology creates gene clusters using correlation-based clustering, which are then used to write association rules that are applied to testing data to filter out the required gene sequences. Finally, in a large dataset, SVM is used as a classification method to identify the class labels of the test gene sequence.
Keywords:	IoT, SVM Classification, Correlation clustering, Gene Sequencing
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	MILLENNIAL PARENTS' PERCEPTION OF PARENTING STYLE THROUGH INSTAGRAM AND WHATSAPP SOCIAL MEDIA IN INDONESIA
Author:	TANIYA INDRIANA RUSTAM, F. FEBRIYANTI BASTARI, CICI FAKHRUNNISA SOFYAN, MUHAMAD ARAS
Abstract:	This study aims to determine the perceptions of millennial parents regarding parenting style obtained through Instagram and WhatsApp social media. Social media is a place to find information about parenting styles, which can then be applied within the family. The type of research is a explanatory research using an quantitative approach. The analysis technique is a quantitative verification analysis through structured interviews using a questionnaire. The population in this study are followers of an Instagram account and members of the parenting group on WhatsApp. The type of sample used in this study is Lemeshow. Data analysis in this study uses the Partial Least Square (PLS) approach. The results of this study are all variables have a positive and significant effect. In this study, there are two related variables, which are X1 (authoritative parenting style) and X2 (authoritarian parenting style). Therefore, it can be concluded that information about parenting styles obtained from social media positively impacts parents' perceptions of parenting style.
Keywords:	Social Media, Millennial Parents, Parenting Style, Authoritative, Authoritarian.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	SECURITY MODEL FOR CLOUD SERVICES BASED ON A QUANTITATIVE GOVERNANCE MODELLING APPROACH
Author:	SWAPNA DONEPUDI , MADHURI A ,V SHARIFF, V KRISHNA PRATAP, S PHANI PRAVEEN, NGUYEN HA HUY CUONG
Abstract:	The security of the cloud, in whatever form it may take, is the most important issue to consider. This necessitates the development of efficient security measurable evaluation techniques for the purpose of shielding data, services, and infrastructure from assaults that may be carried out. The cloud is now receiving a lot of attention in the market, but most businesses are not yet prepared to move their operations to the cloud for the simple reason that safety is their primary worry. No of the nature of the service being utilised by the customer, the business as well as the service provider have the responsibility for maintaining the system's security. As a direct consequence of this, a paradigm for the study of system simulation constituting governance has been provided for cloud-based systems. This article presents a cloud asset mapping and quantifiable governance security evaluation model. Components of this model include asset classification, evaluation, mapping of an appropriate security model, followed by security scanning, a security repair model, and a security quantifiable governance evaluation model. This security elevation model includes a set of assessment aspects that are relevant to a variety of areas, such as networking, maintenance, security application development, and computing, amongst others. The user G-Cloud platform is essential for the successful implementation of the quantitative governance evaluation for many cloud users. This solution walks users through the process of enhancing operation, altering configuration, and finding vulnerabilities using visual graphs to present a dynamic scanning security score. The ranking may include one or many clouds. Doing things right in order to make the cloud's resources safer. This security evaluation system protects the virtual assets of the business as well as the physical organisation by giving better security solutions.
Keywords:	Asset Mapping, Asset Classification, Cloud Asset, Cloud Security, , Quantifiable Governance Security Evaluation.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	BURGLARY DETECTION IN THE RESIDENTIAL AREAS USING YOLO AND SSD ONE-STAGE DETECTION ALGORITHM: A COMPARATIVE PERSPECTIVE
Author:	PAVITHRA.S , Dr.B. MURUGANANTHAM
Abstract:	Deep learning facts are becoming increasingly important and popular. It is a significant factor in object detection. Moreover, object detection, one of the most critical and challenging parts of deep learning, is mainly used in people's everyday lives in areas like video surveillance and self-driving cars. This research is to identify and detect burglars in residential areas. To get a complete and deep understanding of the primary development status of the object (burglary) detection system, we started by looking at the existing algorithms for detection models and setting up our dataset. This paper explains how single-shot multibox detectors and Yolo use deep-learning algorithms to find burglars in residential areas. This research evaluates the deep learning approach for cutting-edge burglary detection systems. This research paper compares two algorithms of one-stage object detection, i.e., you only look once, and single-shot multibox detectors to determine which algorithm is the most effective and fastest. In this comparison, we used the datasets for each algorithm to look at how well they worked, where they were weak, and where they were strong. Based on the results, it says that in the testing environment, Yolo does better than SSD.
Keywords:	Yolo, SSD, Burglary Detection, Deep Learning, Residential
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	A REVIEW OF GAMIFICATION TOOLS TO BOOST STUDENTS’ MOTIVATION AND ENGAGEMENT
Author:	CHE KU NURAINI CHE KU MOHD, SITI NURUL MAHFUZAH MOHAMAD, HAMZAH ASYRANI SULAIMAN, FAAIZAH SHAHBODIN, NUR RAIDAH RAHIM, AIZAD AIZUDIN
Abstract:	Students today are growing up in an increasingly technology-driven world and are becoming more accustomed to using technology. They were grown using digital tools. Significant issues in adapting the learning process to students with various learning preferences must be addressed by educators. One of the educational strategies and methods that increase students' motivation and engagement is gamification. The purpose of the paper is to present and discuss the nature advantages of gamification and provide some suggestions on how to implement it in teaching and learning. In this paper, authors have listed the gamification tools that can be implemented during teaching and learning. This is important to assist the transformation of teaching and learning in higher education and preparing students for future employment in a dynamic environment highly influenced by technology and digital trends. The primary findings could be utilized as guidelines or a resource for gamification solutions to help educators and students engaged in structured learning.
Keywords:	Digital Tools, Online Learning, Gamification, Game Learning, Higher Education
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	INBOUND AND OUTBOUND INTERNET APPLICATION SERVICES QOS ANALYSIS ON LAN METRO-E NETWORK
Author:	NOR PAEZAH ABDULLAH, MURIZAH KASSIM,, YUSNANI MOHD YUSSOFF, SAYANG MOHD DENI, ABDUL JALIL RADMAN
Abstract:	This paper presents an analysis of internet application services utilization on the Local Area Network (LAN) Metro-E network. A recent problem identified that the internet line has been upgraded frequently, but still, the campus faced internet link congestions time by time. The Quality of Services (QoS) on the internet LAN faces the same problem. This study presents the performance analysis of internet application services on LAN Metro-E Network. The data was collected using the Exinda Network Orchestrator, which was set to monitor and control the network throughput. Six months of data were collected on the campus Metro-E network at the main gateway to the internet. The result presents a classification of the top 10 internet application categories used in the network. Inbound and outbound data were analyzed based on the maximum and minimum throughput on the network. The top three inbound application services are others, YouTube and Steam applications, respectively at 70.82%, 20.88%, and 1.94%. Meanwhile, the top three outbound services are others, iCloud and TCP 902-52640 respectively at 82.86%, 5.63% and 5.10%. The Pareto distribution model was identified for both inbound and outbound internet LAN Metro-E network traffic analysis. This research is significant in deploying traffic scheduling, policing, and shaping future QoS on the LAN Metro-E Network.
Keywords:	Internet Application Services, Performance Analysis, Throughput, Quality of Services (QoS), LAN Metro-E, data network
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	AUTOMATED CHANGE IMPACT ANALYSIS TOOL FOR SOFTWARE MAINTENANCE PHASE
Author:	P JALAJA , T ADILAKSHMI
Abstract:	Often, software projects are enhanced based on clients' requests or companies' policies. This process happens in the maintenance phase of the project. The maintenance phase of the Software Development Life Cycle (SDLC) occurs after the product is in production. Maintenance of software can include software upgrades, repairs, and fixes. The required changes are mentioned in the form of change request (CR) which is given to the developers who calculate the impact of changes on the software application. The effect must be assessed without modifying the software application. The change impact analysis is one such way that helps in assessing the impact. It takes time for the developer to analyze the change request and the software application to identify the required changes. In addition, impacts of the changes need to be identified, which requires more time for the developer. And also there is a high chance of missing the impacts due to manual effort. Rather, we can use an automated tool that does the change impact analysis within less time and give more accurate results. The objective of this paper is to develop and test the automated CIA Tool (StaticPy) that performs the change impact analysis. With the help of the tool developer will have an idea of the changes to be made. This will reduce the cost and time. So, this paper proposes a Change Impact Analysis Tool (CIAT) - StaticPy that helps to identify the impacted files, methods, fields, and elements that are affected because of the proposed changes. CIAT takes a change request and project repository GitHub link as input. Then the tool does static analysis on the given repository and forms a data structure. This data structure contains all the details of the impacted elements. From the data structure, we display the information of affected files, methods, fields, and impacted line numbers in the files. StaticPy has been developed and tested using four different software applications. The average accuracy of the tool is 97.8%.
Keywords:	Change Impact Analysis, CIA Tool, data structure, static analysis, Tokenizer.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	ASSESSING CUSTOMER SATISFACTION OF CUSTOMER RELATIONSHIP MANAGEMENT SYSTEM IN AN E-COMMERCE COMPANY IN INDONESIA
Author:	ROBERT MARCHELINO TAJIB , SFENRIANTO
Abstract:	This research aims to explore the customer satisfaction of customer relationship management system (CRMS) in Tokopedia, Indonesia. The growth of e-commerce in Indonesia is big, with Tokopedia as the top e-commerce company with largest visitors. However, there are still many customers that does not feel satisfied with their services which customers to move to other e-commerce platforms and downgrade the ratings of Tokopedia. The variables used in this study are perceived ease of use, perceived usefulness, system quality, user support, and social influence. A sample of 113 respondents of Tokopedia’s customers that lives in Jakarta was gathered from online questionnaire. This study uses the partial least square–structural equation modeling (PLS-SEM) method to analyze the collected data. The SmartPLS 4 4.0.8.7 software was used to perform the data analysis. Based on the research, perceive ease of use, system quality, and user support influence the customer satisfaction, while perceive of usefulness and social influence do not influence the customer satisfaction.
Keywords:	CRM, Customer Satisfaction, E-commerce, System Quality, User Support.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	CONDITIONAL-AWARE SEQUENTIAL TEXT GENERATION IN KNOWLEDGE-ENHANCED CONVERSATIONAL RECOMMENDATION SYSTEM
Author:	K YOGESWARA RAO , K SRINIVASA RAO ,S VENKATA SURYA NARAYANA
Abstract:	Conversational Recommender System (CRS) has become an emerging application that enables user interactions to determine their products or services. An integrated dialog and recommender system design in the CRS necessitates further development in the text generation process to suggest a richer set of textual content during interaction without compromising the recommendation quality. The existing CRS research solutions neglect the provision of the proper guidance to the user until they obtain suggestions for their desired products or services. Even though multi-turn conversational recommendation models guide the users by utilizing their static and dynamic preferences, modeling a recommender system with the desired response as the recommendations is challenging. In addition, providing a recommendation at fewer turns is significant to avoid too many irrelevant turns during the conversation. Hence, the dialog system needs to precisely generate the preference-aware response utterances for a successful recommendation within a minimal set of turns. Thus, this work aims at enhancing the dialog system by associating a Conditional-Aware decision-making process in CRS, namely, CACRS that ensures an optimal transition between the non-recommendation case, referring to the reply utterance generation and recommendation case, referring to the recommendation list generation. The proposed CRS utilizes both the static or historical and dynamic conversations and the external Knowledge Graph (KG) to extract the user preferences by the recommender system and generate the response utterances by dialog system. In the proposed system, the dialog module is responsible for generating the response utterances by the sequential deep learning model and enforcing the optimal transition in the multi-turn conversation by contemplating the preference correlation between the static and dynamic conversation-based recommendations. Thus, the proposed system accurately suggests the final response utterance tagged with the recommended product or service to the users in the CRS through the knowledge-guided recommender system and dialog system.
Keywords:	Recommendation System, Dialog System, Conditional-Aware, Decision-Making, Knowledge Graph, And Optimal Transition.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	A COMBINED AHP-TOPSIS MODEL FOR THE EVALUATION AND SELECTION OF TRUCK DRIVERS
Author:	IMANE BENALLOU , ABDELLAH AZMANI , MONIR AZMANI
Abstract:	Accidents involving heavy trucks result in severe human and material damage. This severity is mainly due to the weight and difficulty controlling the truck. Human error is often the cause of road accidents, hence the interest in carefully choosing the appropriate driver to deliver an order. In fact, the drivers likely to deliver an order must be evaluated according to a set of criteria to choose the one with the least risk of causing an accident. In solving selection problems, multi-criteria decision support methods are often used in most domains. In this paper, we address the problem of decision-making in the transportation domain and, more precisely, the driver selection problem. We propose a model based on two multi-criteria decision support methods, AHP (Analytic Hierarchy Process) and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), to select the driver who has the most negligible probability of causing an accident. A driver's choice is justified by ranking all the candidate drivers according to a range of criteria using the combined AHP-TOPSIS model. The AHP method determines and calculates the relative weights of the decision criteria, while the TOPSIS method is used to obtain the final ranking of alternatives. The prioritization of the evaluation criteria was done based on brainstorming with experts in the field, which allowed us to provide a decision support tool for carriers to evaluate their drivers before assigning them to different routes. The results indicate that the use of medicinal products containing Gemfibrozil and Glibenclamide and the driver's affliction with diabetes are the main criteria in the driver selection process.
Keywords:	AHP, TOPSIS, Truck Driver Selection, Decision Making, Multi Criteria Decision Method.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	SPECTRUM SHARING USING SOCIAL SPIDER OPTIMIZATION BASED STACKELBERG GAME THEORY
Author:	ANOOJA.B, B. BALATRIPURA SUNDARI, P.SUPRIYA
Abstract:	The spectrum sharing technique aids telecom operators in sharing the electromagnetic spectrum by allowing unlicensed users to use the licensed user spectrum. By this, the operators can continuously provide services to users in 4G or 5G networks without any intervention. Here, Stackelberg's game theory (SBGT) is used to facilitate spectrum sharing between the primary and secondary users based on the concept of Game Theory. It provides a better solution with the help of social spider optimization (SSO) by optimizing the cost and band usage. The proposed method is implemented in the frequency range between 800 MHz and 1800 MHz which is used for high-speed 4G and 5G connections. With the use of SBGT, the performance is higher i.e. the efficiency of the spectrum sharing and allocation is higher than others. The basic spectrum sharing is performed after windowed kurtosis spectrum sensing using a threshold method that determines the idle spectrum in the band. Users with low arrival rates affect the performance of a network. Herein we use the SSO method to mitigate the users with low arrival rates to avoid interference and then to determine the rewards of band utilization SSO is combined with SBGT. By using this optimized game theory approach, the users of the licensed band and unlicensed band with low arrival rates are awarded minimum rewards and are avoided for a better quality of service.
Keywords:	Spectrum Sensing, Spectrum Sharing, Windowed Kurtosis, Social Spider Optimization, Game Theory.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	DISCOVERY OF PARENTS' INTENTION FACTORS IN THE USE OF PARENTAL CONTROL ON THE GADGETS USED BY THEIR CHILDREN
Author:	PAULUS IMAN, RIYANTO JAYADI
Abstract:	Children interact with technology and the internet every day by using gadgets such as smartphones, tablets, game consoles, or computers for education or entertainment, but unfortunately, children are prone to be addicted to gadgets. Parental control offers a technical mediation strategy for parents to supervise gadget use. A quantitative study was conducted by distributing a closed-answer questionnaire to parents living in the Jakarta Metropolitan Area with children under age 15 to discover what intentions can influence them to use parental control by proposing some hypothesized factors from parents and children. The number of participating respondents was 423 and the responses were analyzed using Smart PLS by assessing validity, reliability, and hypotheses testing. The study reveals that effort expectancy and parent’s awareness are strong predictors of intention to use parental control. Parents perceive that gadget addiction is related to online risks, health risks, and academic concern in children. The total effect shows that gadget addiction is significantly related to the intention to use. Parent’s age is a strong moderation predictor of self-awareness and parents' self-efficacy to the intention to use. Three parents were interviewed to seek their opinion on why they had not or rarely used parental control even though they knew about it. This study fills the gap between parental control use among parents and gadget addiction in children with potential risks related to online, health, and academic performance.
Keywords:	Parental Mediation; Technical Restriction; Smartphone Addiction; Internet Addiction; Gadget Use By Children.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	A NEW WEIGHTED FEATURE SELECTION AND ENSEMBLE LEARNING WITH HYBRID HEURISTIC STRATEGY FOR DEEP VEIN THROMBOSIS PREDICTION
Author:	KAPARTHI UDAY , Dr.MUDIGONDA MALINI
Abstract:	Deep Vein Thrombosis (DVT) is caused due to blood clots in deep veins. DVT symptoms differ from patient to patient. The prediction of DVT is essential in the primary stage. Otherwise, it may lead to Pulmonary Embolism(PE). In this research, a new DVT prediction model is designed with the data augmentation concept, a technique commonly used in machine learning to enhance the dataset used in the learning process. It produces new cases from the original dataset without varying the pattern of the data. The proposed model consists of data augmentation, weighted feature selection, and classification phases. Initially, the data augmentation is performed after gathering the data related to DVT. Secondly, data cleaning and outlier removal are done on the collected data to improve the data quality. Furthermore, weighted feature extraction is used with Hybrid Lion-Sea Lion Optimization (HL-SLnO) to extract high-quality, non-redundant features from data. Finally, the obtained features are sent to the classification phase, where they are processed using Modified Ensemble Classifiers (MEC) with Long Short-Term Memory (LSTM), Deep Neural Network (DNN), and XGBoost with parameter optimization derived from the same HL-SLnO. MEC combines the decisions from the classifiers mentioned above to improve the overall performance to obtain the best-predicted result. Finally, the experimental results show that the proposed model has effectively enhanced the prediction of DVT risk and provided decision support for other nursing and medical intervention.
Keywords:	Deep Vein Thrombosis, Weighted Feature Selection, Hybrid Lion-Sea Lion Optimization, Deep Neural Network, Long short-term memory.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text

Title:	AN IMPROVED OPTIMIZATION BASED CLASSIFICATION MODEL FOR SENSOR BASED HUMAN ACTIVITY RECOGNITION
Author:	S.SELVABHARATHI , Dr. K.S.DHANALAKSHMI , P.PRABHU
Abstract:	Human Activity Recognition (HAR) is the field of study which infer insights for human activities. HAR plays vital role in various real-world applications includes checking regularly for a period of time about reducing the risk in daily living activities in healthcare systems, elder care and rehabilitation. There are plenty of conventional algorithms exists for HAR and still needs improvement in various factors such as privacy and accuracy. HAR can be classified as vision based such as surveillance video/image, and sensor based such as smartphones, smart watches and wearable devices. The vision based HAR use external devices to collect the data with issues such as privacy, data storage size, cost, infrastructure and accuracy. The sensor based HAR uses on-body smart devices in the wrist, chest, abdomen, ankle, leg and bell to collect the data. Sensors embedded in these devices provides privacy-aware relevant data with information for analysis. This research work aims to consider all aspects of sensor based human activity recognition, datasets, pre-processing techniques, optimization techniques, classification models and prediction accuracy by proposing Random Forest(RF) and Support Vector Machine Classifier (SVMC)based GSCV-RF and GSCV-SVM methods. These methods outperform in terms of prediction accuracy when compared with conventional models.
Keywords:	Classification, Human Activity Recognition, Machine Learning, Optimization, Prediction, Sensors.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2023 -- Vol. 101. No. 7-- 2023
Full Text