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

Welcome to Journal of Theoretical and Applied Information Technology

Submit Paper / Call for Papers

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

Journal of Theoretical and Applied Information Technology
March 2019 | Vol. 97 No.07

Title:	DEEP NEURAL CLASSIFICATION AND LOGIT REGRESSION BASED ENERGY EFFICIENT ROUTING IN WIRELESS SENSOR NETWORK
Author:	Mrs. J.SRIMATHI,AP(SG)/MCA, Dr. B. SRINIVASAN
Abstract:	In Wireless Sensor Network (WSN), routing strategies are required for distributing the data from sensor nodes to base station. During the data transmission, the node energy is the key parameter for improving the network lifetime. The conventional routing techniques are developed to perform routing in WSN but it failed to lessen the energy consumption and improve the network lifetime with minimum overhead. In order to overcome the above-said issue, Energy-Efficient Deep Neural Node Classifier based Logit Regressed Routing (EEDNC-LRR) technique is introduced. The main aim of EEDNC-LRR technique is to perform energy efficient routing and increase the reliability of data transmission with maximum network lifetime and minimal overhead. In EEDNC-LRR technique, the sensor nodes are taken to transmit the environmental data. Initially, all sensor nodes have a certain energy level. A sensor node consumes some amount of energy during the sensing of data in WSN. The Deep Neural Node Classifier model is used in EEDNC-LRR technique to classify the node as higher energy nodes and lesser energy nodes based on threshold energy level. From that, the deep neural node classifier model produces the efficient classification results of sensor nodes. This helps to choose the higher energy nodes to reduce the energy consumption and extend the network lifetime while routing the data packets. Subsequently, the higher energy nodes are transmitted to the output layer for efficient routing in WSN. With higher energy nodes, Logit Regression Analysis is carried out to identify the nearest neighbor node through a time of arrival (ToA) to identify the distance between the source nodes and sink node. After selecting the nearest neighbor, the route with a minimum distance between the nodes is discovered for routing in WSN. This in turns, the reliable data packets transmission is achieved with minimum overhead. The simulation is conducted with various parameters namely energy consumption, network lifetime, reliability and routing overhead with respect to a number of sensor nodes and data packets. The simulation results and discussion shows that EEDNC-LRR technique improves the network lifetime, relaibility and also minimize the routing overhead as well as energy consumption.
Keywords:	WSN, Routing, Deep Neural Node Classifier, Threshold Energy Level, Logit Regression Analysis, Nearest Neighbor Node, Time Of Arrival.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	AN INTELLIGENT HOME ENERGY MANAGEMENT (IHEM) BASED ON STATE OF CHARGE OF BATTERY IN HOUSEHOLD LOADS ON RENEWABLE ENERGY SYSTEM
Author:	OYINKANOLA LATEEF O. ADEWALE , SAWAL HAMID BIN MD ALLI , RAMIZI MOHAMED, AREMU OLAOSEBIKAN AKANNI
Abstract:	This work proposed the switching control strategy for household appliances in a rural area that based their storage on the battery for three days. The system is to track the state of charge (SOC) of the battery to balance with the demand from the output appliances. The existing controller or regulator tries to off all the appliances when the SOC is reaching a certain level. However, to avoid a total blackout and to avoid at least an appliance to be working. Thus, we try to design an intelligent discrete-time controller and optimize the controller parameter for the set point control followed the base rule. The system performance result shows that proposed switch control strategies work within a certain limit and avoid total discharge of the battery.
Keywords:	Alternative energy, Battery, Controller, Intelligent, and SOC
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	DYNAMICALLY CONFIGURABLE MANET TOPOLOGY FOR WIRELESS NETWORKS USING DCNRPT ALGORITHM
Author:	Mrs. J.VIJAYALAKSHMI, Dr. K.PRABU
Abstract:	Mobile Ad hoc Network (MANET) is a collection of mobile nodes that communicate between themselves on wireless links and are different from wireless LAN (WLAN) communications. MANET’s mobile nodes are end devices routing information. Mobility of nodes increases near geographical edges of the network clusters. The nodes in a MANET work as both a sender and receiver of information. The basic qualities of reconfiguration and ease in deployment of MANETs make them a suitable candidate for emergency applications. MANET’s network topology is based on the relative positions of connecting nodes. Links get created and break when nodes change positions within the network. Node mobility affects source, destination and intermediate nodes, resulting in an extremely volatile topology. The dynamism of MANET topology is a unique challenge and hence a dynamic topology based on ad-hoc position of Mobile nodes is proposed in this paper and also improved packet delivery ratio, throughput, jitter and reduced delay time. The proposed topology DCNRPT is simulated in NS2 for deployment of nodes and an optimized coverage.
Keywords:	MANET, DCNRPT, WCA, Topology Based Routing, Position Based Routing, Topology Characteristics.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	A PRIORITY-QUEUE DOMINATED AND DELAY-ENERGY GUARANTEED MAC SUPERFRAME STRUCTURE FOR WBANS TO DEAL WITH VARIOUS EMERGENCY TRAFFIC OF PILGRIMS DURING HAJJ: AN ANALYTICAL MODEL
Author:	SHAH MURTAZA RASHID AL MASUD, ASMIDAR ABU BAKAR, SALMAN YUSSOF
Abstract:	In recent times, IEEE 802.1.5.6 based wireless body sensor networks (WBANs) are being deployed in various medical and healthcare centres for providing quick and real-time health facilities among the patients. Every year during Hajj several millions of pilgrims gather together at overcrowded ritual site ‘Kaaba’ and its surrounding places in Makkah. Ensuring the best healthcare facilities and services among the pilgrims those who are suffering from a variety of critical conditions and illness including chronic diseases; sudden illness, trauma and accidents at the congested environment is a demanding research issue because lack of proper healthcare facilities may worsen the life of pilgrims. In our research, we define the emergency medical data into five different criticality levels which are aperiodic or random and require immediate transmission to the healthcare stations for further actions. Based on the data criticality level a priority-criticality index table and a modified superframe structure for medium access control (MAC) protocol are developed. Critical or emergency data is obligatory to be transmitted ahead of other non-critical traffic as delay in its transmission may impede human life. But, the problem may occur when more than one emergency data from different sensors aggregate to the coordinator for further transmission to the healthcare stations through exclusive access period-EAP slot of MAC. For smooth, low delay and energy efficient data transmission we propose an analytical method based on M/M/1 priority-queuing system to provide delay-energy guaranteed QoS at MAC level in our research. The Poisson distribution process has been used to analyse the packet arrival rate of each queue. In this research, M/M/1 queuing system is used to investigate five different priority-queues for five emergency traffics. Emergency traffic is defined and prioritised through the criticality level of pilgrims’ medical problems as being generated by body sensors. From the mathematical analysis, we have seen that the medical data with the highest priority should not stay at the queue for a long time which decreases the queuing delay at the system. Moreover, for energy efficient data transmission, we propose a sleep/idle-wakeup mechanism that reduces unnecessary energy consumption. An extensive analytical approach through mathematical model shows the better performance comparing to the default IEEE 802.15.6 standard which assumes the sensor nodes are awake all the time. Finally, in our future research, extensive experimental work will be conducted to corroborate and validate the analytical findings which are mentioned in the discussion and future work section.
Keywords:	Hajj, Pilgrims, Emergency, WBANs MAC Superframe, Low Delay, Energy Efficiency, Queue
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	BEHAVIOR ANALYSIS OF THE USE OF E-LEARNING USING UTAUT MODEL APPROACH (CASE STUDY: STMIK MIKROSKIL)
Author:	ZULPA SALSABILA, EDI ABDURACHMAN, SOPHYA HADINI MARPAUNG
Abstract:	E-learning has become one of the factors that needed by universities to be able to compete and survive. Electronic learning (e-learning) using Internet and digital technologies to create experience in educating others. E-Learning in STMIK Mikroskil used to help students and lecturers in teaching and learning process. This study uses the UTAUT (Unified Theory of Acceptance and Use of Technology) model. The aim is to analyze the tendency of users of e-learning systems at STMIK Mikroskil Medan, by testing whether Behavioral Intention and Behavior to Use a technology are influenced by Performance Expectancy, Effort Expectancy, Social Influence, and Facilitating Conditions. These four factors are moderated by gender, experience, and voluntary factors. Questionnaire data were collected from 346 active students and analyzed by structural equation modeling (SEM) using AMOS 24. The results of this study showed Performance Expectation Factors, Effort Expectation Factors, Social Influence Factors had a positive effect on Behavioral Intention, Facilitating Conditions Factors had a positive effect on Use Behavior, Behavioral Intention has a significant influence on Use Behavior. Gender does not have a moderating effect that affects the factors of Performance Expectation and Effort Expectancy on Behavioral Intention, but Gender has a moderating effect that affects Social Influence factors towards Behavioral Intention. Experience does not have a moderating effect that influences Business Expectancy factors on Behavioral Intention. However, experience has a moderating effect that influences Social Influence factors that have a positive effect on Behavioral Intention. Experience also has a moderating effect that affects the Facilitating Conditions factor for Use Behavior. Voluntariness has a moderating effect that affects Social Influence factors towards Behavioral Intention.
Keywords:	E-Learning, Performance Expectancy, Effort Expectancy, Social Influence, Facilitating Conditions, Unified Theory of Acceptance and Use of Technology (UTAUT)
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	A SECURE CLOUD-BASED PICTURE ARCHIVING AND COMMUNICATION SYSTEM FOR DEVELOPING COUNTRIES
Author:	ADEBAYO OMOTOSHO, JINMISAYO ADIGUN AWOKOLA, JUSTICE ONO EMUOYIBOFARHE, CHRISTOPH MEINEL
Abstract:	Picture Archiving and Communication Systems (PACS) are used to enable medical images from imaging modalities to be stored electronically and viewed on screens so that medical practitioners and other health professionals can access them. However, PACS comes with a lot of costs for storage, air conditioning licenses and so on, these necessitate the need to take advantage of existing technologies that can enhance adoption, especially in developing countries. More alarming is that data centres for electronic health services have been the target of several attacks and hacks in recent year, even though cloud computing has the potential of making the adoption of PACS more cost-effective. Cloud computing has a major drawback in the area of security. In this work, a framework for securing cloud-based PACS is developed and implemented.
Keywords:	E-Health, Telemedicine, PACS, Cloud Computing, Security, Modalities, Medical Images, Privacy
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	PERFORMANCE ANALYSIS ON THE ARDUINO UNO MICROCONTROLLER-BASED WEIGHT MEASUREMENT SYSTEM FOR TODDLER
Author:	KURNIAWAN TEGUH MARTONO, EKO DIDIK WIDIANTO, YUSUF BAHCTIAR
Abstract:	Error in reading the data or in processing the data input frequently occurs. It might be caused by the user negligence, the condition of the device used, or the unstandardized system and this, later on, leads to the problem in terms of data accuracy. The computer technology development today makes it possible to help human in coping with the problems mentioned above. In measuring the weight of the under-five-year-old children, the errors in reading and inputting data can occur. This is in relation to various abilities of each cadre. Also, the use of the device to measure the weight is not in accordance with the standard – some might use the electronic weighing scale and other might use the mechanical one. This unstandardized device can emerge a problem - particularly in reading the data. The weighing measurement system using Arduino is one of the ways to control the growth of the under-five-year-old children and this can minimize the error level in reading and inputting data. The development of this system uses the sequential development method and the application of this model can result in the best result. The test of the system was conducted through two phases: phase 1 related to the test conducted in laboratory and phase 2 related to the test conducted in Posyandu (maternal and child health centre). The results of the tests showed the error percentage of 0.067% and the percentage of accuracy of 99.93%. The result for the time taken in sending the data from the system to the mobile communication device was at 1.97 seconds at distance of 1 meter and 2.24 seconds at the distance of five meters.
Keywords:	Measurement, Arduino, Performance, Data, Weight
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	AUTOMATIC SECURITY EVALUATION OF SPN- STRUCTURED BLOCK CIPHER AGAINST RELATED-KEY DIFFERENTIAL ATTACKS USING MIXED INTEGER LINEAR PROGRAMMING
Author:	HASSAN MANSUR HUSSIEN, SHARIFAH MD YASIN AND, ZAITON MUDA, NUR IZURA UDZIR
Abstract:	Block cipher algorithms become an essential domain in Information Technology (IT) due to ever increasing the number of attacks. In point of fact, it is significant to produce a security evaluation of block cipher algorithms to determine a statistical non-random behavior of attacks. In relation to this, a new theoretical attack such as related-key differential cryptanalysis (RDC) could give rise to a more practical technique. Basically, estimating immunity of lower bounds in the substitution-permutation network (SPN) block ciphers structure against RDC attack is essential for providing a secure block cipher algorithm. Currently, the automatic computer tools are not applicable to estimate the immunity against related-key differential attacks for SPN block ciphers structure. We present a searching strategy that determines the lower bounds of SPN block ciphers structure against RDC using the Mixed Integer Linear Programming (MILP). This study also aims to demonstrate the applicability and the efficiency of the MILP technique by examining the security of Rijndael block cipher in RDC attack. We prove this technique through calculate the number of activation S-boxes into Rijndael block cipher. The extended MILP technique is able to provide an automatic security estimation tool by giving accurate results. Overall, it is applicable to an extensive variety of block cipher algorithm that makes it an adaptable tool for industrial purposes and scholarly research.
Keywords:	Related-key Differential Cryptanalysis, Mixed Integer Linear Programming (MILP), SPN- structured Block Cipher, Rijndael, and Automatic Search Tool
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	EVALUATING THE ACCURACY OF DATA MINING ALGORITHMS FOR DETECTING FAKE FACEBOOK PROFILES USING RAPIDMINER, WEKA, AND ORANGE
Author:	MOHAMMED BASIL ALBAYATI1, AHMAD MOUSA ALTAMIMI, DIAA MOHAMMED ULIYAN
Abstract:	Facebook is on constant growing, attracting more users due to the provided high-quality services for Online Socializing, Sharing Information, Communication, and alike. Facebook manages data for billions of people and is therefore be a target for attacking. As a result, sophisticated ways for infiltrating and threatening this platform have been developed. Fake profiles, for instance, are created for malicious purposes such as financial fraud, impersonate identities, Spamming, etc. Numerous studies have investigated the possibility of detecting fake profiles on Facebook with each study focusing on introducing a new set of features and employing different machine learning algorithms for countermeasure. This paper adopts a set of features from the previous studies and introduces additional features to improve classification performance in order to detect fake profiles. The performance of five supervised algorithms (Decision Tree, Support Vector Machine SVM, Naïve Bayes, Random Forrest, and k Nearest Neighbour k-NN) are evaluated across three of the common mining tools (RapidMiner, WEKA, and Orange). The experimental results showed that SVM, Naïve Bayes, and Random Forest had a stable performance with a nearly identical results across the three mining tools. However, Decision Tree outperformed other classifiers on RapidMiner and WEKA with accuracy of 0.9888 and 0.9827, respectively. Finally, we observed that k-NN showed the most significant change with an accuracy of 0.9603 for WEKA, 0.9145 for Orange, and 0.9460 for RapidMiner tool. These findings would be useful for researchers willing to develop a machine learning model to detect malicious activities on social network.
Keywords:	Fake Profiles Detection, Machine Learning, Classification, WEKA, RapidMiner, Orange
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	APPLICATIONS OF MACHINE LEARNING ALGORITHMS TO THE PROBLEM OF DETECTING UNKNOWN DATA
Author:	AKHMER YERASSYL, AKHMER YERMEK, BEKTEMYSSOVA GULNARA UMITKULOVNA, USKENBAYEVA RAISA KABIEVNA
Abstract:	Nowadays, the term big data means the work with large volumes and diverse composition, very often updated and located in different sources in order to increase work efficiency, create new products and increase. The problem of understanding a large number of unstructured and previously unknown data for the task is usually manually solved by a team of mathematicians and analysts who find it difficult to retain the value of all the data and all the hidden relationships of the incoming multidimensional array. However, in order to classify data, a full understanding of it is required. Today, there can be a lot of data sources: data from sensors of critical equipment (“Internet of things”), transactional “tires” and databases, electronic documents and paper media. In this regard, the quality of classifiers of most of the studying models decreases significantly. This article describes various machine learning methods for the classification of previously unknown data and ways to improve the quality of models. The advantages and disadvantages for each of the models are described, as well as the necessary and sufficient conditions for use. As experimental data, the Otto Product Classification dataset were taken from the open database.
Keywords:	Machine Learning, E-Commerce, Classification, Decision Trees
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	PREDICTION OF OSTEOARTHRITIS USING LINIER VECTOR QUANTIZATION BASED TEXTURE FEATURE
Author:	LILIK ANIFAH, MAURIDHI HERY PURNOMO, TATI L R MENGKO
Abstract:	Osteoarthritis estimated as the eighth-leading nonfatal burden of disease in the world is the one important reason why it is investigated. The status of osteoarthritis is important because it is used as a basis for determining treatment for patients. The aim of this research is analyzing the texture feature of Junction Space Area (JSA) and design system based texture feature in order to predict the severity of osteoarthritis in the knee using a linear vector quantization. Textures extracted in this study are first order, second order, and gray level run length matrix. Several stages involved as the research procedures covering image processing, feature extraction, learning process, and testing process. The result of feature extraction obtained several FO, GLCM and GLRLM features for each cluster with overlapping conditions, making it difficult to classify using linear methods, so learning used linear vector quantization (LVQ). Feature extraction was carried out for both training data and testing data. The training process, which was divided into several stages namely first order learning, GLCM learning data, GLRLM learning data, and combined learning data features. learning process for the aforementioned features of combined learning data used learning parameter rate of 0.5 with epoch values of 1000, 5000, 10000, and 15000. The best results obtained when using a system using LVQ based on GLCM features. But the disadvantage of this system is that it cannot recognize grade 2 well where recognizing grade 2.
Keywords:	Osteoarthritis, FO, GLCM, GLRLM, LVQ
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	OPEN AND CLOSED EYES CLASSIFICATION IN DIFFERENT LIGHTING CONDITIONS USING NEW CONVOLUTION NEURAL NETWORKS ARCHITECTURE
Author:	NOOR D. Al-SHAKARCHY, ISRAA HADI ALI
Abstract:	A classification of eyes status (open and closed) is a most important step in various application, such as drivers fatigue detection, psychological states analysis. Monitoring a driver to detect inattention is one of the important application can be achieved base on the period that the eyes staying closed. Therefor the core of these systems is the perfect classification of the status of eyes open or closed. Traditional methods to classify open and closed eyes suffer from the luminance information. These systems performed inadequately if images of eyes obtained with various resolutions and elimination conditions. Most computer vision applications deal with images in different lighting conditions and resolutions as well as the real life applications must take into account the accuracy and time of execution. Deep Neural Network presented the efficient method to extract robust feature such as eye features, which determine the status of eyes closed or open. The proposed method for classification the eyes status which those open or closed in images of eyes with various elimination conditions based on new architecture of Deep Neural Networks (DNNs). The proposed Deep Neural Network Classification (DNNC) performed accuracy with reasonable training and simulation time. Performance speed is the most crucial point in real-time applications when the training and performance time of existing methods have long times. The primary goals of the proposed (DNNC) can be achieved the less time spend for execution as well as this system can be implemented in reasonable hardware capability. The proposed system can achieve accuracy with training set reach to 96%. The values of loss function reach to 0.01.
Keywords:	Deep Neural Network (DNN), Convolution Layer, Maxpooling Layer, Relu Activation Function, Sigmoid Activation Function
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	THE APPLICATION OF DATA MINING METHODS FOR THE PROCESS OF DIAGNOSING DISEASES
Author:	SAULE BELGINOVA , INDIRA UVALIYEVA , SAMIR RUSTAMOV
Abstract:	Today, the medical field has a large amount of medical information that requires proper processing and further use in the diagnosis and treatment of various diseases. The development of computer technology provides tremendous opportunities for collecting, processing, managing and researching medical information to better understand the complex biological processes of life and help solve the problem of diagnosis and treatment in medical institutions. Accurate diagnosis and proper treatment provided to patients is one of the main tasks of medical care. Therefore, data mining techniques, which are part of knowledge discovery in databases, are becoming popular tools for medical researchers. The use of these tools makes it possible to identify and use patterns and relationships between numerous variables and to predict specific disease and treatment outcomes. This paper describes the application of data mining methods for the process of diagnosing diseases. To improve the effectiveness of the methods of data mining for the process of diagnosing diseases, a procedure for assessing the informativeness of heterogeneous diagnostic indicators on the basis of a theoretical information approach has been proposed. The results of diagnosing are described on the basis of revealing the relationship between the supply of magnesium and the risk of somatic diseases with the use of intelligent data analysis.
Keywords:	Diagnosis, Information Signs, Intellectual Analysis Of Medical Data, Diagnostically Valuable Signs, Clustering.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	A COMPARISON OF CHANGE MANAGEMENT GUIDELINES TO ADDRESS TECHNOLOGY ADOPTION BARRIERS: A CASE STUDY OF HIGHER EDUCATIONAL INSTITUTIONS
Author:	HABIB ULLAH KHAN , RICHARD GARETH SMUTS
Abstract:	Emerging technologies are bolstering the development in the world in all the domains of the world. Education sector is an example of such innovation, by basing on which new means of pedagogy and andragogy are being explored with suitable models. The present study aims to know the difference in such models and the change management guidelines in addressing the barriers of Technology Enhanced Learning (TEL) in Higher Educational Institutions (HEI). The students of the Department of Electrical, Electronic and Computer Engineering (DEECE) of 1st year are selected from the Cape Peninsula University of Technology (CPUT) are selected for the study using purposive random sampling. They are divided into two groups and are made to participate in a survey after providing some level of orientation about the change management models and the role of the model such as UTAUT and ADKAR. The opinion of the students is collected regarding the ability of the models to hedge against the barriers in the process of technology enhanced learning. The collected data is analyzed to check the means and standard deviation of scores and hence the difference between the opinion of the groups. To achieve this, t-test for difference of means with unequal variance is used. The results led to the acceptance of the null hypotheses that are framed basing on the research questions. Among the two groups of students surveyed, it is proved that the group that got exposed to TEL approach is able to understand the barriers in a better manner than their counterparts. In addition to the results, the draw backs of the study like limited sample size, clarity about the causality and ineffective participation of students are identified. Hence it is advocated to the future research works to overcome the limitations and to understand the role of the change management guidelines to address the barriers in the technology adopting as per the setting.
Keywords:	Technology Enhanced Learning (TEL); Areas of Tension (AOT); Unified Theory of Acceptance and Use of Technology (UTAUT); Adoption of Learning Technology (ALT); Awareness/Desire/Knowledge/Ability/Reinforcement (ADKAR); Higher Education Institution (HEI)
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	EXTRACTING PATTERN FROM LARGE DATA CORPUS USING APPROXIMATE REASONING AND INTUITIONISTIC FUZZY CLUSTERING TECHNIQUES
Author:	Dr. ASHIT KUMAR DUTTA
Abstract:	The fuzzy logic is a familiar method to automate a complex activity. Intuitionistic fuzzy clustering is the extension of fuzzy logic. Approximate reasoning is a concept to deal vague and complex data. Many real time problems were solved with the help of Intuitionistic fuzzy concepts. Classification and clustering are the well known methods in the process of extraction of knowledge from the large data corpus. The existing techniques that are based on clustering methods could not provide optimum results with less computation cost. The efficiency of the existing methods was not up to the mark on large datasets. The objective of the proposed research is to provide an efficient technique for the extraction of meaningful pattern from large dataset with least computation cost. The proposed method has used the Intuitionistic fuzzy clustering technique with approximate reasoning for the extraction of pattern from the benchmark datasets. The experiment and results on large data corpus has proved that the level of the proposed research is satisfactory.
Keywords:	Fuzzy clustering, Pattern extraction, Pattern Miner, Fuzzy – C Means, Soft computing
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	A NEW PERMUTATION METHOD FOR SEQUENCE OF ORDER 28
Author:	ABDULLAH AZIZ LAFTA, AMMAR KHALEEL ABDULSADAH, SAFAA JASIM MOSA
Abstract:	Permutation is reordering for a set of objects with taking into consideration the important of order of their locations. In order to increase security performance, this paper presents a new method of permutation for sequence of order 2^8. In this propose method, any information can be converted to the sequence of approximate values between 0 and 255 as a set of blocks. It is needed to build the so called lookup table so that it contains the frequency and positions of the original sequence values. In the experiments, twenty-six image files have been experimented as evidence to the proposed method. In addition, the correlation and entropy measures are computed to test the quality of permutation. All the tests observed that method have significantly effective in reducing the correlation and thereby decreasing the perceptual information of the sequence. Hence, the security is improved.
Keywords:	Permutation, Algorithm, Frequency, Sequence, Byte
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	AN EFFICIENT TECHNIQUE FOR CLUSTER NUMBER PREDICTION IN GRAPH CLUSTERING USING NULLITY OF LAPLACIAN MATRIX
Author:	IMELDA ATASTINA, BENHARD SITOHANG, G.A.PUTRI SAPTAWATI, VERONICA S.MOERTINI
Abstract:	Clustering graph dataset representing users’ interactions can be used to detect groups or communities. Many existing graph clustering algorithms require an initial cluster number. The closer the initial cluster numbers to the real or final ones, the faster the algorithm will converge. Hence, finding the right initial cluster number is important for increasing the efficiency of the algorithms. This research proposes a novel technique for computing the initial cluster number using the nullity of the Laplacian Matrix of Adjacency Matrix. The fact that nullity relates to the properties of the eigenvalues in the Laplacian matrix of a connected component is used to predict the best cluster numbers. By using this technique, trial and error experiments for finding the right clusters is no longer needed. The experiment results using artificial and real dataset and modularity values (for measuring the clusters quality) showed that our proposed technique is efficient in finding initial cluster numbers, which is also the real best cluster numbers.
Keywords:	Estimating the Number of Clusters, Nullity, Laplacian Matrix, Adjacency Matrix, Graph Clustering
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	THE ONTOLOGY APPROACH FOR INFORMATION RETRIEVAL IN LEARNING DOCUMENTS
Author:	LASMEDI AFUAN, AHMAD ASHARI, YOHANES SUYANTO
Abstract:	The number of documents on the Internet has increased exponentially. Every day, users upload various documents to the Internet. This raises a problem, how to find content documents that are relevant to user queries. Information Retrieval (IR) become a useful thing to retrieve documents. However, IR still uses a keyword-based approach to content search that has limitations in displaying the meaning of the content. Often, keywords are used mismatch and miss concept with a collection of documents. As a result, IR displays documents that are not relevant to the context of the information needed. To overcome these limitations, this study has applied Ontology-based IR. The dataset used in the study is 100 learning documents in the field of Informatics which include lecture material, practicum modules, lecturer presentations, proceedings articles, and journals. IR performance evaluation is done by comparing ontology-based IR with classical IR (keyword based). We evaluate IR performance by executing ten queries for testing. Documents that retrieves by query execution are calculated for performance by using Precision, Recall, and F-Measure evaluation metrics. Based on IR performance evaluation, obtained average recall, precision and f-measure values for ontology-based IR of 88.11%, 83.38%, and 85.49%. Meanwhile, IR classics obtained average recall, precision, and f -measure 78.70%, 70.96%, and 74.47%. Based on the values of Recall, Precision, and F-Measure, it can be concluded that the use of ontology can improve relevance document.
Keywords:	Information Retrieval, Ontology, Learning Document, Precision, Recall, F-Measure.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	RT-VC: AN EFFICIENT REAL-TIME VEHICLE COUNTING APPROACH
Author:	SALAH ALGHYALINE, NIDHAL KAMEL TAHA EL-OMARI, RAED M. AL-KHATIB, HESHAM Y. AL-KHARBSHH
Abstract:	This paper proposes and implements an efficient real-time vehicle counting (RT-VC) approach. This approach is based on the most efficient action detection and tracking methods in computer vision. YOLO is used for object detection, whereas Kalman filter with Hungarian algorithm are used for tracking. The road is divided into two zones of interest by the end-user, and any vehicle will be counted if its trajectory crosses these zones of interest. The experiments show that the proposed system is very accurate in comparison with other existing approaches. For comparative evaluation, our proposed approach obtained accuracy above 90% for most of the tested videos in the highway roads. Therefore, the proposed approach can efficiently work with many real-time surveillance systems and has the potential to be used in many real road applications.
Keywords:	Artificial Neural Networks (ANNs), Convolutional Neural Network (CNN), Artificial Intelligence (AI), Deep Learning Algorithms, Vehicle Counting, Surveillance Systems, Traffic Managements.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	CLUSTER TRACE ANALYSIS FOR PERFORMANCE ENHANCEMENT IN CLOUD COMPUTING ENVIRONMENTS
Author:	HAYDER H. MAALA , SUHAD A. YOUSIF
Abstract:	Cloud computing has received considerable interest from research institutions, developers, and individuals in the last years. A trace’s cluster of approximately 12,500 machines, referred to as the "Google cluster trace” had been initiated by Google. This paper examines the characteristics, download process and tools, and analysis of this trace dataset in an attempt to provide insights into a type of trace date similar to the date which is in the cloud environment. We analyzed trace dataset by using the K-means clustering algorithm executed over SQL Server to use the implemented methodology for enhancing cloud environment performance by allocating the data into clusters. This allocation was aimed to be used in distributing the upcoming tasks to the most suitable cluster, then to the most suitable machine which covers its need for resources. The clustering process generates some clusters depending on CUP rate for each task, these clusters represent the machines suitable to each range (Average) of CPU rate which is required from the upcoming task to be allocated. Depending on the relationship between tasks and machine data, machines could be selected of each produced cluster for calculating the availability of CPU usage. This calculation will be the millstone of future tasks allocation over cloud cluster machines depending on its resources availability and its suitability to future task resource requirements.
Keywords:	Google Cluster Trace; Clustering; Performance Enhancement; K-means; Cloud Computing;
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	REVIEWS ON EDGE-BUNDLING TECHNIQUES FOR PARALLEL COORDINATES
Author:	N. NADIA S. ASRI, ZAINURA IDRUS, SITI ZZ ABIDIN, H. ZAINUDDIN, MT MISHAN
Abstract:	Parallel coordinates are well-known visualization tool for high and multi-dimensional datasets. However, data cluttering and over plotting data are major issues for parallel coordinates and becoming worst in the era of Big Data. Thus, edge-bundling technique has overcome the aforementioned problems by bundling similar edges in parallel coordinates. This paper provides in-depth reviews on six approaches of edge-bundling techniques which are cost-based approach, geometry-based approach, image-based approach, force-directed edge-bundling approach, divided edge-bundling approach and lastly, hierarchical edge-bundling approach. On top of that, these particular approaches are able to reduce the visual clutter without changing the positions of the nodes and only the shapes of edges have some changes. Based on the findings, each approach supports certain features of parallel coordinates such as shape, scale, weight, position, direction and length. This paper suggests a relationship diagram for all these features. The relationship model was proposed in order to improve data visualization by removing data clutter. In summation, edge-bundling techniques are applicable to parallel coordinates in order to visualize multivariate datasets.
Keywords:	Parallel Coordinates, Edge-bundling, Data Visualization, Information Visualize, Big Data
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	FACTORS AFFECTING ONLINE CONSUMER BEHAVIOR ON THE LEADING E-MARKETPLACE IN INDONESIA
Author:	SUTJAHYO, TOGAR ALAM NAPITUPULU
Abstract:	The research objective is to analyze the factors that influence online consumer behavior in e-marketplace in Indonesia and its priorities. Questionnaires were distributed to 384 respondents who were users of one of the top five e-marketplaces in Indonesia. Data were analyzed using Structural Equation Modeling (SEM) and processed using SmartPLS. This research model describes 80.50% of online consumer behavior, where there are several factors, namely seller's reputation, trust in market-maker, perceived ease of use, perceived usefulness, perceived risk, after sales services, customer services, trust in the seller, and product quality. It can be concluded that the seller's reputation factor is a factor that must be prioritized in increasing online consumer behavior.
Keywords:	Online Consumer Behavior, E-Marketplace, SEM, SmartPLS, Indonesia
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	DEEP BELIEF NETWORK BASED QUESTION ANSWERING SYSTEM USING ALTERNATE SKIP-N GRAM MODEL AND NEGATIVE SAMPLING APPROACHES
Author:	K YOGESWARA RAO, GORTI SATYANARYANA MURTY, T.PANDURANGA VITAL
Abstract:	The Question Answering (QA) system becomes essential owing to the increasing amount of web content and high demand for the right and short information. Intending to enhance QA results towards the Natural Language Processing (NLP) community, most of the Question answering system exploits machine learning algorithms to generate an appropriate answer to the user query. Even though, it lacks to predict accurately over the large-scale data by itself, needs an external force to make adjustments in the answer prediction. With the recent evolution in deep learning, the neural network architecture reflects its potential for QA. The deep learning model can determine the issues in answer prediction on their own and resolves it. The class of deep neural network such as Deep Belief Network (DBN) is widely applied in question answering especially for text processing. Moreover, most of the works on the text processing exploits the skip-gram model for representing the relevant words in the vector over a massive volume of unstructured text data. However, it results in inefficient outcomes, especially when processing the combination of frequent data and stop words. To resolve these issues, this paper introduces the Deep Neural network for Answering user queries (DNA). The proposed DNA approach performs the QA system over DBN by applying alternate skip-N gram and negative sampling. The conditional probability measurement develops an alternate skip-N gram model and alternatively applying the normal N-gram and skip-N gram model. It improves the efficiency of relevant word-pair detection without increasing the computational complexity. By only using samples, the negative sampling reduces the impact of noise on the accuracy of alternate skip-N gram model and improves the efficiency of the QA system. Finally, the DNA is evaluated using Java and compared with the existing Unified model for Document-Based Question Answering (UDBQA). The results show the efficiency of DNA, for instance, the UDBQA approach reduces the F-measure by 16.3%, compared to the DNA approach with 2000 number of queries.
Keywords:	Deep Belief Network, Question Answering, Skip-N Gram Model, Negative Sampling, And Hidden Layers.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text

Title:	ALTERNATING SENSING PROCESS TO PROLONG THE LIFETIME OF WIRELESS SENSOR NETWORKS
Author:	MOHAMMED AL-SHALABI, MOHAMMED ANBAR, ALAA OBEIDAT
Abstract:	In the last few decades, the usage of Wireless Sensor Networks (WSNs) is dramatically increased due to the nature of this type of networks where the nodes are randomly deployed in a large area to sense some attributes which the human beings cannot do so and send the data to the main device called Base Station (BS), and this process takes a time interval called round. The deployed nodes may be located near to each other and then they may sense approximately the same events. The process of sensing and sending the same data to the BS consumes a lot of energy in the nearby nodes and they will die fast. An alternating sensing process is proposed. The goal of the proposed mechanism is to prolong the lifetime of the network by increasing the lifetime of the nearby nodes and reducing the consumed energy in them. This goal is achieved by scheduling the sensing process to reduce the consumed energy as much as possible. The MATLAB tool is used to evaluate the proposed mechanism and compare it with the LEACH protocol. The results show that the proposed mechanism outperforms the LEACH protocol in terms of the network lifetime and the number of transmitted packets to the BS by approximately 9% and 17% alternatively.
Keywords:	Wireless Sensor Networks, Network Lifetime, Sensing Process, Nearby Nodes, Leach Protocol.
Source:	Journal of Theoretical and Applied Information Technology 15^th April 2019 -- Vol. 97. No. 07 -- 2019
Full Text