		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
November 2020 | Vol. 98 No.22

Title:	INFLUENTIAL ELEMENTS OF SELECTION OUTSOURCING PROVIDER IN IT GOVERNANCE
Author:	GLORY UREKWERE ORLU, NORAINI CHE PA, RODZIAH BINTI ATAN, HAZLINA BT HAMDAN
Abstract:	Over the years, decision outsourcing IT Provider selection in the software industries is rapidly spreading globally. Based on this, organizations outsource their non-core project (software development) to the IT Provider in order to achieve their strategic goals. However, organizations still face challenges in outsourcing provider selection process which lead them in making mistakes of choosing inappropriate IT Provider for their stipulated project. This paper investigates on the challenge faced in outsourcing IT Provider selection process and presents selected influential elements that can support IT practitioners. This study considered three sections during the investigation of related elements for outsourcing provider selection process. In section 1, thirty elements were identified, hence were screened to ascertain their justification from prior researches and suitability in this study. In section 2, based on the screened process undertaken, eleven elements were compared together in order to ascertain the level of preceding researches support towards elements for outsourcing provider in IT governance services, thereafter, in section 3, only six influential elements were considered to be highly supported by prior researches and this was taken for justification in this study. Since human judgment is imprecise and vague, the application of the selected influential elements considered in this study serves in solving the uncertainty and complexity in decision outsourcing provider selection process.
Keywords:	Decision-making; Outsourcing; Provider Selection Practices; IT Governance.
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	ENERGY-LEAKS IN ANDROID APPLICATION DEVELOPMENT: PERSPECTIVE AND CHALLENGES
Author:	MUHAMMAD UMAIR KHAN, SHANZA ABBAS, SCOTT UK-JIN LEE, ASAD ABBAS
Abstract:	Number of mobile devices are increasing, and popularity of Android OS supported devices is more than other, and Android apps dominated the mobile apps market. These devices have limited resources (CPU, Memory, and Power, etc.). Energy consumption in mobile devices is an important factor to consider when developing an app. There is no such official guide that help developer to build a less energy-hungry app or discover energy leaks in the development phase. This paper will review different types of energy leaks in Android apps, how these leaks effect the energy consumption of the device. We will discuss how these energy leaks can be avoided at the development phase
Keywords:	Energy Leaks, Code Smell, Android Apps, Energy Consumption, Mobile Apps, Program Analysis
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	DEEP LEARNING FOR DOCUMENT CLUSTERING: A SURVEY, TAXONOMY AND RESEARCH TREND
Author:	LUBNA GABRALLA, HARUNA CHIROMA
Abstract:	Information is stored in several forms such as pictures, web pages, sound and video, but 80% is stored as a text. Quick searching for a specific text document depends totally on the accuracy of the classification of the document's subject with a similar group of documents. This process is called documents clustering. Recently, deep learning techniques have achieved distinguish results in solving the problems facing documents clustering such as complex semantics and high dimensionality. This paper aims to examines a comprehensive review related to documents clustering, and survey the recent work in document clustering using deep learning methods. The proposed taxonomy represents knowledge that helps researchers to understand and follow up previous works in this area, and developing or creating new methods and a comparative analysis was made between popular dataset, performance metrics, deep learning frameworks and library used in deep learning clustering documents.
Keywords:	Clustering, Deep Learning, Documents, Texts
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	OPTIFOG: OPTIMIZATION OF HETEROGENEOUS FOG COMPUTING FOR QOS IN HEALTH CARE
Author:	PRATIK KANANI, DR. MAMTA PADOLE
Abstract:	A patients life can be saved if it is possible to make quicker decisions based on faster processing of real-time health care data, such as ECG processing. To achieve faster decision making, contemporary health care applications use cloud computing for such data. When cloud computing is used, data transmission deferrals may cause delays in the decision-making process. To overcome this, Fog Computing is used. Fog Computing saves energy, bandwidth and prevents transmission latencies but, lacks in computing power as compared to Cloud Computing. To enhance the computing power of the Fog node, a Cluster of Raspberry Pi having heterogeneous configurations can be used. In Health Care applications the Fog Computing performance can be assessed by measuring the time elapsed between the generation of the health care data and decision-making. In this paper, ECG signal analysis is taken as a processing job in Fog Computing. Dispy is used to facilitate the scalability and parallel data processing on a Cluster of Raspberry Pi used for Fog Computing, to enable faster decision making. Further, the performance of the Raspberry Pi cluster-systems using dispy are analyzed and optimized step by step based on different parameters. The first parameter is data transmission time which is improvised by minimizing network overheads. Other optimization parameters like CPU usage, number of cores, response time and available memory space, these parameters are considered and varied, to assess the performance of Heterogeneous Raspberry Pi cluster. Based on the results obtained, a novel optimization approach “OptiFog” is proposed to achieve faster computation in worst-case scenarios by varying and assigning jobs to the nodes to measure performance parameters in Distributed Fog Computing. Based on the obtained results “OptiFog” assures best possible improvement in the performance of the Distributed Fog Computing environment.
Keywords:	Dispy, Distributed Fog Computing, ECG Analytics, Heterogeneous Fog Computing, Optifog, Optimization In Fog Computing, Raspberry Pi Cluster
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	DEVELOPMENT OF INDONESIAN TALKING-HEAD ANIMATIONS ON HUMANOID AND NON-HUMANOID CHARACTERS USING THE RETARGETING PROCESS OF FACIAL MOTION CAPTURE DATA
Author:	ARIPIN, MUHAMMAD NASRULLOH
Abstract:	The difference in the facial structure of the human and the 3D virtual model (Non-Humanoid) is a challenge in the talking-head animation development. This difference resulted in the talking-head animation of the 3D virtual model that does not match with the talking-head animation of the human models. One example is the difference in the mouth width of a human model and the 3D virtual model. This research aims to develop the talking-head animation on several 3D virtual characters based on the facial motion capture (MoCap) data. The facial MoCap data are recorded using MoCap technology. A radial basis function (RBF) method is used to process retargeting of facial MoCap data on several 3D virtual characters. This method is used to map the feature points from the face source to the target face. The experimental results showed that the talking-head animation on several 3D virtual models can imitate the mouth movement of an actor. The result of the space transformation on tortoise face model has a standard deviation of 0.028261. The value of a relatively small standard deviation shows that the talking-head animation on a tortoise face model corresponds to the mouth movement of an actor. We also evaluate the talking-head animation quality on 3D virtual characters using MOS (Mean Opinion Score) method. The result of MOS calculation shows that the talking-head animation on several 3D virtual characters is 4.133. It means that the 3D virtual characters can imitate mouth movements of an actor.
Keywords:	Talking-Head Animation, Retargeting Process, 3D Virtual Characters, Facial MoCap Data, Radial Basis Function
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	FACTORS ASSOCIATED WITH THE ADOPTION OF MOBILE-BLENDED LEARNING TO DEVELOP CRITICAL THINKING SKILLS IN NIGERIAN BUSINESS EDUCATION UNDERGRADUATE STUDENTS
Author:	AUGUSTINE AGBI, SUPANEE SENGSRI
Abstract:	The standard of the Nigerian educational system has continued to witness a decline over the years at all levels, contributing to the high rate of unemployment among graduates in the country. Jobs in the 21st century require employees’ abilities to analyse, evaluate, and synthesize knowledge to creatively solve problems and these abilities are increasingly necessary for survival in the modern world. Mobile-blended learning allows sufficient time during its face-to-face components to effectively engage learners with problem-solving tasks that encourage their cognitive development, rather than exposing them to learning experiences that are fraught with a repetitive accumulation of facts and knowledge only. The main objective of this study was to investigate the factors associated with the effective adoption of mobile-blended learning for critical thinking enhancement. A cross-sectional study that used survey and documentary research was conducted among 120 business educators from three states of the thirty-six states of Nigeria, A validated self-administered questionnaire was used to collect data, which were analysed using multiple regression method. The study revealed that teachers’ participation in the process of deciding to adopt innovation, competency, mobile instructional content, and mobile-blended learning orientation were significantly associated with effective adoption. This suggests that when adopting a mobile-blended learning approach, the focus should be centred on teachers’ participation and competency as well as mobile content and orientation.
Keywords:	Mobile-Blended Learning, Critical Thinking Skills, Problem-Solving
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	PIPELINE AND DEEP LEARNING APPROACH FOR NLIDB: A COMPARATIVE STUDY
Author:	SHANZA ABBAS, MUHAMMAD UMAIR KHAN, SCOTT UK-JIN LEE, ASAD ABBAS
Abstract:	Databases are integral part of current world’s scenario of rich technology. Greater amount of the data in the world is stored in the databases. That amount of data storages can be utilized for various purposes in data science world. Besides potential usage and benefits of available data amounts, the requirement of formal language to access the databases is a huge hurdle. Structured Query language (SQL) is one of such formal languages to access the database. Besides its impact and powerful as a language it is not a common knowledge. Therefore, domain experts of some particular databases cannot access their data freely and easily. Web interfaces to access that data has their own limitation and do not fulfil the purpose to the maximum of the potential of data. Natural Language Interface to Database (NLIDB) system is natural solution for such problems. Text to SQL task in NLIDB system is being experimented with since 70s. Previously it was based on integrated methods and techniques from Natural Language Processing (NLP) and Data Science areas, those integrated frameworks generally known as pipeline methods. Recently, machine learning showed promising performance for the solutions to semantic problems. Which is why, deep learning had been adopted for text to SQL task as well. Currently NLIDB systems research is going on with both of the approaches of pipeline methods and deep learning methods in parallel. It is important at this time to analyze the latest work with both approaches and compare and identify their unique challenges and issues as well as findings and potential of both approaches for the NLIDB systems. In this paper, a comparative analysis is presented to find out the achievements and issues of NLIDB with pipeline methods and with deep learning methods regarding each of them.
Keywords:	Structured Query language, Natural Language Processing, Natural Language Interface to Databas
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	A PROPOSED MODEL TO PREDICT AUTO INSURANCE CLAIMS USING MACHINE LEARNING TECHNIQUES
Author:	SHADY ABDELHADI, KHALED ELBAHNASY, MOHAMED ABDELSALAM
Abstract:	One of the main challenges facing the insurance companies is to determine the proper insurance premium for each risk represented by customers. Risk differs widely from clients to another, and a Careful understanding of various risk factors assists predict the likelihood of insurance claims based on historical data, Real-world datasets often have missing values, can cause bias in results. the most widely adopted methods for dealing with missing data is to remove observations having missing values, perform a complete case analysis (CCA) and single imputation such as average. these approaches have the disadvantages represent in loss of precision and biased. The main objective of the paper is to build a precise model to predict car insurance claims through machine learning techniques. with a focus on advanced statistical methods and machine learning algorithms that are the most suitable method for handling missing values. we Used available datasets through Kaggle which consists of 12 variables and 30240 cases, the research was carried out by using Artificial Neural Network (ANN), Decision Tree (DT), Naïve Bayes classifiers, and XGBoost to develop the prediction model. The experimental results showed that the model obtained acceptable results The XGBoost model and Resolution Tree achieved the best accuracy among the four models, with an accuracy of 92.53% and 92.22%, respectively.
Keywords:	Machine Learning; Prediction Model; Missing Data; Auto Insurance Claims
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	SEMANTIC REPRESENTATION OF MUSIC DATABASE USING NEW ONTOLOGY-BASED SYSTEM
Author:	AHMED M. DAOUD , KHALID M. HOSNY, EHAB R. MOHAMED
Abstract:	The Semantic Web is suffering from a lack of tools used to facilitate its users’ work when extracting structured data. The development of a system that facilitates the extraction of data from databases and converts them into useful information is of great interest. In this paper, a new ontology-based method is proposed to extract the structured data from a database based on a target ontology. The main focus here is on dealing with music databases because of their popularity on the web. Since the Semantic Web built upon reusing existing ontologies, this system is going to use the existing musical ontology defined in the Music Ontology Specification. This specification provides the concepts and properties used to describe music besides other ontologies that could use in conjunction with it, such as (foaf, dc, timeline, event, etc.). The resulting music data, in RDF format, can be published and linked with existing musical data on the web. This method contributes to the music web of data and allows the Semantic Web clients to access detailed structured information about musical data easily.
Keywords:	Semantic Web (SW), Relational Database (RDBs), Resource Description Framework (RDF), Music Ontology, Music Database
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	FEATURE SELECTION AND CLASSIFICATION OF SPEECH DATASET FOR GENDER IDENTIFICATION: A MACHINE LEARNING APPROACH
Author:	RIZWAN REHMAN, KAUSTUVMONI BORDOLOI, KANKANA DUTTA3, NOMI BORAH, PRIYAKHI MAHANTA
Abstract:	In speech analysis, gender identification is one of the most complex tasks. Gender can be traced from the acoustic parameters like formants (F1, F2, F3, F4) or the pitch (F0). Therefore it is very important to identify which feature or features can classify the dataset efficiently in terms of a male and female speakers. This paper is an attempt to classify the dataset more accurately using fewer features i.e. among F0, F1, F2, F3, and F4. For the feature selection, the Fisher score algorithm is used to find out the most discriminative feature that can be used for the classification of the gender from the speech data set. Then to cross-validate the result obtained using the Fisher score algorithm we have applied the Tree-based algorithm. The results of both the algorithm comply with each other as F0 or pitch is the most distinctive feature among all with both the algorithms. Since the result of both the algorithm comply with each other we have then performed the classification by applying logistic regression, KNN classifier, SVM, and Decision tree algorithms. We have then evaluated and compared the accuracy of each of the features using these classification techniques. The finding of this study will provide the statistical means to identify the best feature for gender identification from the acoustic characteristic.
Keywords:	Feature Selection, Classification, Gender Identification, Statistical Methods
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	AN INTEGRATED APPROACH OF DYNAMIC TASK SCHEDULING OF DAG WITH DUAL MODE PROCESSORS-USING MACHINE LEARNING TO OBTAIN OPTIMAL MAKE SPAN
Author:	PRASANT SINGH YADAV, P.K YADAV, SUNIL BHARTI
Abstract:	With increasing computing demand the need for tuned intelligence-based solutions is most required. Most of the focus has been given by the researcher to the scheduling of parallel tasks dynamically to more than one processor and in the current scenario, it is more demandable. Although many DAG scheduling algorithms are available but less focused on dynamic scheduling. Through our projected paper we want to introduce the approach Dynamic task scheduling algorithm DTSA for scheduling task at run time using DAG with an additional factor regarding processor self-Reconfiguration Capacity, which is an important parameter of distributed computing System. Through DTSA we want to sketch out an adaptive task arrangement algorithm that gives the hybrid result of run-time scheduling of DAG and adaptation of tenant configuration by the processor according to computing needs. Finally, A DAG-based dynamic task arrangement with dependency consideration between the tasks and with the use of machine learning (ML) for self-reconfiguration of a processor is proposed for obtaining the optimal task allocations with the optimal Makespan.
Keywords:	DAG, DTSA, LTA, TPC-W, CPU Self-Reconfiguration, Machine Learning.
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	AN EXPERIMENTAL STUDY: PERSONALIZED GAMIFED LEARNING BASED ON LEARNING STYLE
Author:	RASHA ALJABALI, NORASNITA AHMAD, AHMAD FADHIL YUSOF, SURAYA MISKON, NAZMONA MAT ALI, SALAMATU MUSA
Abstract:	Numerous game related elements such as points, badges, leaderboard, and ranking are utilized in educational context when motivation and engagement need to be boosted. Undoubtedly, influencing students’ engagement has been reported as one of the gamification’s pros. However, study the effects of certain game elements on dissimilar types of students is recommended by several researchers. Accordingly, a learning style is a vital factor in human learning process which is considered as an important personalization parameter in several eLearning tools. This research has used Felder-Silverman Learning Style Model (FSLSM) as personalization parameter along with 10 game elements to propose a personalized gamified learning model. A Design Science Research Approach (DSRA) has been undertaken to examine the proposed model in improving the students’ scores in Data Flow Diagram (DFD) lesson during class learning process. For validating the proposed model, 50 Multiple Choice Questions for DFD web-based gamified application has been developed. An experimental study using the application has been conducted with 71 undergraduate students from School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia (UTM). Participants were divided into two groups: experimental and control. Additionally, the gamification application has two different modes: personalized mode (for experimental group) and non-personalized mode (for control group). Data was collected from the application database and perceived usefulness questionnaire. An independent t-test has been used to compare means of scores between the groups. Result shows that personalized gamified learning is an effective method in learning process, as well as in boosting student perceived usefulness of the application.
Keywords:	Gamification, Learning Style, Personalized Learning
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	ONTOLOGY-BASED SEMANTIC INTEGRATION OF HETEROGENEOUS DATA SOURCES USING ONTOLOGY MAPPING APPROACH
Author:	KITTIPHONG SENGLOILUEAN, ROMCHAT KHUNTONG
Abstract:	At present, educational institutions have developed their library database systems, or used library package programs. Some educational institutions had many libraries such as the university libraries, the faculty libraries, and the department libraries. However, standards in developing and designing the library systems were differently applied. These differences caused the problems when sharing the link of data from many libraries. That was, it encountered the problem of centralized semantic query because each library stored the same data, but chose to use different vocabulary to describe the data. As there was no common agreement or standard for describing data, it brought about the semantic conflicts of synonyms and homonyms. In this paper, the ontology-based semantic integration of heterogeneous data sources by using the ontology mapping approach was implemented for solving such the problem. The results of the study revealed that the architecture of the semantic integration of heterogeneous data sources using the ontology mapping approach could effectively resolve the semantic conflicts of linking the library data from various sources. Based on the performance evaluation of the proposed architecture, it was found to be highly effective with Precision, Recall, and F-measure, at 90.34%, 91.98%, and 91.15%, respectively.
Keywords:	Ontology Mapping, Semantic Integration, Semantic Conflicts, Library Ontology
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	ANALYZING FACTORS AFFECTING SATISFACTION AND PURCHASE INTENTION TOWARDS MOBILE AUGMENTED REALITY E-COMMERCE APPLICATIONS IN INDONESIA
Author:	STEFANIE LIU, TOGAR ALAM NAPITUPULU
Abstract:	The high internet economy value in Indonesia is supported by e-commerce, with beauty products as one of the top spending categories. Nevertheless, the majority of Indonesian women still prefer to buy beauty products at outlets. In 2019, augmented reality feature started to present in mobile e-commerce applications and allows users to try the product virtually. This research aims to analyze factors affecting satisfaction and purchase intention in mobile augmented reality e-commerce applications in Indonesia. Quantitative method was used in this research, with data collection through questionnaires. Data from 403 respondents were processed in SmartPLS 3.0 software. This research revels that purchase intention was affected by satisfaction, trade off price and value, perceived augmentation and perceived enjoyment. Meanwhile satisfaction is affected by trade off price and value, perceived augmentation, perceived usefulness, perceived enjoyment and system quality.
Keywords:	Augmented Reality, E-commerce, M-commerce, Purchase Intention, Satisfaction
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	FACTORS AFFECTING PURCHASE INTENTION ON TIX.ID MOBILE APP
Author:	EDWARD CHANDRA , TOGAR ALAM NAPITUPULU
Abstract:	Internet and smartphone nowadays have evolved, many people make use of it to do online transaction. One of them is to purchase movie ticket through online. TIX.ID itself is a mobile app that have a feature to buy movie tickets online. In this case study, it has a purpose to know factors affecting purchase intention on TIX.ID mobile app. Factors that has been used on this case study are Perceived Ease of Use, Perceived Usefulness, Trust Security, Subjective Norm and using Age as moderating variable to find out whether Age has affect on connection between those variables with purchase intention. In this case study, it uses quantitative method by using questionnaire that has been sent to respondents as data source for this case study. From the questionnaire that has been sent through social media, it received 451 respondents which the respondents have a qualification to ever at least once to purchase movie tickets through TIX.ID mobile app. Result from this case study is that factors affecting purchase intention on TIX.ID mobile app are Perceived Ease of Use, Perceived Usefulness, Security and Subjective Norm.
Keywords:	Online Movie Ticket, E-commerce, M-commerce, Purchase Intention, TIX.ID
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	PERFORMING A CONTENT VALIDITY: ESTABLISHING A RELIABLE INSTRUMENT TO MEASURE THE INTENTION TO ADOPT CLOUD COMPUTING SOFTWARE AS A SERVICE IN PUBLIC ORGANISATION
Author:	HIBA JASIM HADI, MOHD ADAN OMAR, WAN ROZAINI SHEIK OSMAN, MOHAMMED FADHIL IBRAHIM, MUKTAR HUSSAINI
Abstract:	Nowadays, cloud computing software as a service (CC-SaaS) has gained widespread popularity and vast advantages in the information technology domain. However, the adoption rates of CC-SaaS among organisations in developing countries are inadequate and still not widely adopted. Many public organizations are still lacking a broader understanding of adopting and utilizing CC-SaaS to facilitate tasks and increase efficiency. This trend is developing countries is more evident in Iraqi public organisations; thus, it highlighted the need to adopt such technologies to be able to reduce the cost of IT infrastructure and provide fast information accessibility. This paper's main objective is to develop an instrument used to assess the possibilities of CC-SaaS intention to adopt more, especially in Iraqi organisations. Also, to ensure the instrument's validity developed to make sure they are adequate to determine the adoption and avoid the meaningless and uninterpretable experimentation result for the intention to adopt CC-SaaS in a public organization. The paper also describes a systematic approach to assess the research instrument by employing a content validity index for the proposed constructs. A panel of 12 experts was used to validate the instrument through the quantitative (content validity) method, by Item-CVI (I-CVI), Scale-level CVI (S-CVI), and the modified Kappa statistic. The result shows high content validity for the items, and it also helped reduce and modify some of the items. Thus, the results show a high level of trust in the abilities, integrity, and benevolence of CC-SaaS providers will minimize the Iraqi organization's security and privacy concerns and motivate them to acquire the cloud service, as technological, organisational, environmental. Human factors depicted in this paper are valid.
Keywords:	CC-SaaS, Questionnaire, Content Validity, I-CVI, and S-CVI
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	ANALYSIS OF STUDENT LEARNING HABITS FACTOR TOWARDS E-LEARNING EFFECTIVENESS IN XYZ UNIVERSITY
Author:	HANS KRISTIAN, TOGAR ALAM NAPITUPULU
Abstract:	In an overall educational process, learning activities are the most important activities. Along with the times, an innovation is needed in the learning process. E-learning is a learning method that uses electronic media as teaching material for students. At XYZ University, the use of e-learning is also applied in their learning system. The purpose of this study was to determine the effect of student study habits on the effectiveness of e-learning. The data were taken from 157 online student respondents XYZ University who were taking a master program and then analyzed using the smartPLS tool. It is known that the influential variables are Delay Avoidance, Work Method, Self-Efficacy, and Self-Directed Learning. The results of this study indicate that the Work Method and Self-Directed Learning factors have a positive and significant effect on the effectiveness of e-learning and can be considered to increase the effectiveness of the e-learning learning system.
Keywords:	Delay Avoidance, Work Method, Self-Efficacy, Self-Directed Learning, E-Learning Effectiveness
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	SENSING PERFORMANCE ANALYSIS OF AN INTERNET OF THINGS-BASED INTRUDER DETECTION SYSTEM
Author:	ARKOM SUDRAM, WASANA BOONSONG
Abstract:	The Internet of Things (IoT) has become part of human life since it has made interconnection among people and things together. IoT impacts a security because affected an individual’s well-being. This paper contributes a performance analysis of an intruder detection system based on IoT service network platform. The IoT cycle of intruder monitoring/detection system considered are consists of 4 sections, which are 1) intruder monitoring section, 2) communication network, 3) data processing unit and 4) user interface. Each section has its different role but they work together seamlessly and efficiently. The proposed intruder monitoring device system adopted the Ultrasonic and Passive Infrared (PIR) Sensors. Both were tested the moving detection accuracy performance with various distances in detecting a moving object. The findings were analyzed in terms of percentages. NETPIE-IoT cloud server acts as an intermediary to exchange information with other devices on the internet network. The statistical package for the social sciences (SPSS) used for analyzing the experimental results by comparison between the two their performances. The results showed that the Ultrasonic and PIR sensors show different responses based on individual tests. Both intruder sensors have a statistically significant differences in percentages of moving detection accuracy with 99.94 (Std. Deviation = 0.113) and 99.56 (Std. Deviation = 0 .679) respectively. The Ultrasonic sensor was slowly dropped, whilst the PIR sensor was slightly increased according the distance variations.
Keywords:	IoT, Ultrasonic, PIR, NETPIE, SPSS
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	AN OVERVIEW OF DEEP LEARNING TECHNIQUES IN ECHOCARDIOGRAPHY IMAGE SEGMENTATION
Author:	MEHDI SAMIEIYEGANEH, PROFESSOR RAHMITA WIRZA BT O. K. RAHMAT, Dr. FATIMAH BINTI KHALID, Dr. KHAIRUL AZHAR BIN KASMIRAN
Abstract:	Machine Learning (ML) has been a remarkable success in the last few years, Sequential -Decision Making tasks are a main topic in ML, these are tasks based on deciding, the sequence of actions from experience carry out in an environment that is uncertain to achieve goals. these tasks cover so many ranges of applications such as healthcare, robotic, finance and many more. The design and extracting of features in ML were done based on defining (hand -crafting features), which is a weak point for this model. Due to ML problem as well as advances in computer hardware Machine Deep Learning (DL) has entered the field of image processing. In fact, DL is a type of Function approximator. To solve ML tasks Function approximator is required and this idea is core of ML. There are many type of Function approximator such as, linear models, gaussian process, support vector machine and decision tree. In this paper, considering the importance of segmenting in medical images, we will review works that have utilized DL methods, as well as our focus is based on Echocardiography Image Segmentation.
Keywords:	Machine Learning, Deep Learning, Medical Image Segmentation, Echocardiography.
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	REAL-TIME WATER QUALITY MONITORING OF AQUACULTURE POND USING WIRELESS SENSOR NETWORK AND INTERNET OF THINGS
Author:	WASANA BOONSONG1 , WIDAD ISMAIL, NAOKI SHINOHARA, SEVIA MAHDALIZA IDRUS SUTAN NAMEH, SURYANI ALIFAH, KAMARUL HAFIZ KAMALUDIN, TONI ANWAR
Abstract:	Data monitoring with updated information is important and necessary in today’s era of digital technology. This study proposes a smart water quality monitoring (SWQM) system of aquaculture ponds that uses a Wireless Sensor Network (WSN) at a radio frequency band of 900 MHz in an Internet of Things (IoT) platform. The proposed system provides solution in the monitoring and death prevention of aquatic animals in culture ponds. The important parameters studied in this research are temperature, dissolved oxygen in water and potential Hydrogen (pH) values. The information monitored using the proposed SWQM device is transmitted to an operator through a cloud Internet platform via the router gateway. The operator can utilize the tracked information data on a smart device to achieve real-time monitoring. The data are analyzed and further evaluated on the basis of the results of water conditions. This implementation emphasizes the improvement of the agriculture process for the benefit of economic development at the community, society and national levels.
Keywords:	SWQM, WSN, IoT, Dissolved Oxygen, pH
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	UNSUPERVISED KEYWORD EXTRACTION USING NON-SMOOTH NMF
Author:	ALIYA NUGUMANOVA,DARKHAN AHMED-ZAKI, MADINA MANSUROVA, YERZHAN BAIBURIN, KURMASH APAYEV, ALMASBEK MAULIT
Abstract:	In this paper, we introduce a novel unsupervised method for keyword extraction, based on non-smooth nonnegative matrix factorization. We generate a document-term matrix from a given corpus and factorize it into the product of two special matrices: documents-by-topics and topics-by-terms. In our method, we choose a low degree of factorization (k=3,4,5) and use only topics-by-terms matrix to extract top N keywords for each of k topics. Then we merge these obtained N*k keywords into a resulting keyword list excluding duplicates and assign keywords to documents. We validate our method with a large text corpora: “Introduction to information retrieval” textbook (by Manning, Raghavan and Schütze), available online. The result of our method is compared with three popular unsupervised keyword extraction algorithms: TextRank, Rake and Yake. The experiments confirm that the proposed method shows the promising performance in terms of precision, recall and F-measure with respect to various number of candidate keywords.
Keywords:	Keyword Extraction, NMF, nsNMF, NLP, Unsupervised Approach.
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	DETECTION APPROACH BASED ON MULTI-HEAD STRUCTURE AND ENHANCED FEATURES IN DRIVING ENVIRONMENTS
Author:	HOANH NGUYEN
Abstract:	Most recent state-of-the-art detection approaches based on deep convolutional neural networks are manually designed. These approaches include two-stage frameworks and one-stage frameworks. While one-stage frameworks provide real-time performance in most recent systems, two-stage frameworks usually show better detection accuracy. Most recent two-stage object detection frameworks share a head for both classification and bounding box regression in detection stage. Inspired by recent improvement in double-head structures, this paper proposes a detection framework based on multi-head structure for localizing objects in driving environments. First, the extracted feature maps generated by feature extraction network are enhanced by the enhancement module, which effectively enlarges the receptive field and refines the representation ability of thin feature maps by leveraging both local and global context. The enhanced feature map is then fed to a detection network. Next, the detection network is designed based on double-head structure, where a fully connected head is adopted for classification and a convolution head is used for bounding box regression. In addition, this paper proposes to improve RoI pooling algorithm based on deformable RoI pooling. With the improved RoI pooling process, the harsh quantization of RoI pooling is removed, and the extracted features are properly aligning with the input, thus leading to large improvements. Experiments on public datasets show the effectiveness of the proposed method for localizing objects in driving environments.
Keywords:	Detection Approaches, Double-Head Structures, Multi-Head Structure, Deep Convolutional Neural Networks, Two-Stage Framework
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	PARTITIONED GLOBAL ADDRESS SPACE APPROACH FOR THE MAPREDUCE IMPLEMENTATION OF THE PARALLEL KMEANS ALGORITHM
Author:	MANSUROVA MADINA, SHOMANOV ADAY
Abstract:	In recent years there is a growing challenge in processing large amounts of data as the size of the data gets exponentially increasing. Mapreduce became an advanced tool to tackle these problems with processing of large arrays of data. As a result, many current Mapreduce frameworks such as Apache Hadoop, Apache Spark rely on Mapreduce as a backbone technology to solve their large-scale problems. Though, such approaches have their benefits, performance wise they cannot always guarantee a linear speed-up and hence a new parallel methods and frameworks needs a thorough study in order to understand scalability and performance benefits in these cases. In this work we present Kmeans parallel (\|\|) clustering algorithm implemented in a partitioned address space Mapreduce system. This work includes a comparison and performance analysis of the presented implementation. In the paper we propose a novel approach that was not considered in literature before. In particular, it was found that our Mapreduce implementation of Kmeans parallel algorithm achieves a strongly linear speed-up that makes this approach an excellent candidate to solve high-dimensional and large-scale clustering problems.
Keywords:	Mapreduce, PGAS, UPC, K-Means, Clustering
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text

Title:	A FRAMEWORK FOR REAL-TIME DYNAMIC RESCUING SYSTEM FOR INDOOR ENVIRONMENT
Author:	AKEEM OLOWOLAYEMO, SALEH ALANAZI, MOHD SYAFIQ ZAMRI, MITCHELLE LIAW AI WEI, TEDDY MANTORO
Abstract:	Most current emergency operations employ manual find-and-rescue procedures. Consequently, people trapped in a building remain helpless in an emergency, unable to call out successfully until a rescue team come along which often may be too late. The work presented in this study proposes a dynamic real time rescue system approach to an emergency such as fire outbreak in residential multilevel building or apartment. The research focuses indoor environment, even though the idea is adaptable for outdoor environment as well. The study proposes utilizing automated reporting of disaster situation to fire service in the event of a fire outbreak, automated residents’ roll calls to all registered occupants in a building, automated emergency status request push notification to all residents, and dynamic rescue combined with indoor pathway safest route guidance, to guarantee safer rescuing procedures. The dynamic rescue approach employs dynamic trapped resident information mining to deploy firemen proportionately to affected areas. The accuracy of the resident information mining is approximately 97.8 % for large datasets while 90% (9/10) for small datasets. The study proposes strategies to mitigate observed challenges with most of the previous rescuing systems. It is hoped that this study may provide a new direction for emerging smart buildings and future directions for rescuing and emergency situation.
Keywords:	Dynamic Rescuing System, Indoor Safety System, Emergency Response, Indoor navigation, Indoor Dynamic Localization
Source:	Journal of Theoretical and Applied Information Technology 30^th November 2020 -- Vol. 98. No. 22 -- 2020
Full Text