|
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
September 2024 | Vol.
102 No.17 |
Title: |
BLOCKCHAIN TECHNOLOGIES AS A FACTOR OF THE FINANCIAL SUSTAINABILITY MANAGEMENT
OF THE ENTERPRISE AND THE E-COMMERCE DEVELOPMENT |
Author: |
OLHA POPELO, SVITLANA TULCHYNSKA, OLENA ANDRIUSHCHENKO, SVITLANA SHEPELENKO,
MYKYTA FALKO, SERHII SHUT |
Abstract: |
The rapid development of digital technologies fundamentally changes the essence
of business and the direction of its development. This opens up new
opportunities for managing financial stability of enterprises, including through
the blockchain technology, which is also a factor in the e-commerce development.
The purpose of the article is the justification of the feasibility of
implementing blockchain technologies as a factor in managing financial stability
of the enterprise and the e-commerce development is presented. Achieving the
outlined goal using the methodology of the system approach gave the opportunity
to prove the feasibility of implementing blockchain technologies as a factor in
financial sustainability management of the enterprise and the e-commerce
development, which, unlike existing approaches, involves: firstly, taking into
account the stages of the tokenization process; secondly, a methodical approach
to the calculation of costs for the emission and tokenization of securities;
thirdly, determining the economic effect of the emission and tokenization of
securities through the calculation of the difference between the income from the
sale of share tokens and the costs of their creation. Highlighting the varieties
of an alternative approach to improving the level of the financial stability
management of enterprises and the e-commerce development made it possible to
find out the advantages of the asset tokenization model. The use of a systematic
approach enabled the authors to propose a methodical approach for calculating
the costs of the emission and tokenization of securities and determining the
economic effect of the emission and tokenization of securities. Before
approbation of the proposed methodical approach, the prerequisites of
tokenization of securities were analyzed using the example of the institutional
environment of Ukraine and indicators of the business activity of economic
entities. This analysis proved the possibility and expediency of implementing
blockchain technologies as a factor in managing financial stability of
enterprises and the e-commerce development. The approbation of the proposed
methodical approach to determining the efficiency of the additional issue of
shares and their subsequent tokenization was carried out based on the data of
the annual financial statements of the current Ukrainian enterprise Interpipe
NTZ PJSC - an enterprise of the Interpipe Holding, which carries out production
activities. |
Keywords: |
Blockchain Technology, Financial Stability, Enterprise, Business Entities,
E-Commerce, Business, Tokenization, Token, Smart Contract, Fractionalization,
Share Issue, Cryptocurrency, Liquidity. |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
A STUDY OF QUALITY CRITERIA OF MOBILE HEALTH APPLICATION FOR MEDICATION
ADHERENCE: USER VIEWPOINTS |
Author: |
MADIHAH ZAINAL, A. IZUDDIN ZAINAL-ABIDIN, AND SUZIAH SULAIMAN |
Abstract: |
Mobile health (mHealth) applications (apps) have been growing rapidly in the
commercial market aligned with the increase in ownership of smartphones. The
mHealth apps provide diverse functions from various areas including assisting
individuals with complex and long-term treatments to adhere to their prescribed
medications. However, several literatures reported a low number of mHealth apps
that were considered good or high-quality medication adherence apps. Besides,
there was a lack of user involvement in the current quality assessment tool of
mHealth apps for medication adherence. This paper aimed to determine quality
criteria for mHealth apps used for medication adherence based on the potential
users of the apps. A survey was created based on Medication Adherence
Application Quality (MedAd-AppQ) items and several criteria from relevant
literature. The survey was distributed to patients and caregivers of chronic
diseases through physical and online platforms. In total, 207 participants had
completed the survey for this study, with 40% of the participants were chronic
disease patients and 53% of the participants were the caregivers. Based on the
findings, content reliability, feature usefulness, and feature convenience had
positive correlation values, and feature convenience had the highest value with
3.40 (p-value <0.001). This paper revealed diverse quality criteria and
categories that can be integrated into the current quality assessment tool or
the design for mobile health apps. By providing new insights for app developers
and health practitioners, this study hoped to enhance the advancement of current
mHealth apps and eventually improve the health conditions of individuals with
long-term diseases. |
Keywords: |
Medication Adherence, Mobile Health App, Quality Criteria. |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
AUTOMATED WHEAT RUST DISEASE CLASSIFICATION USING DEEP LEARNING TO SUPPORT SDG
1: ENHANCING AGRICULTURAL PRODUCTIVITY AND POVERTY ALLEVIATION |
Author: |
CHBISWARANJAN NANDA, SUDHIR KUMAR MOHAPATRA, RABI NARAYAN SATPATHY, ALIAZAR
DENEKE DEFERISHA, SEIFU DETSO BEJO |
Abstract: |
Agricultural production is pivotal for alleviating extreme poverty and boosting
economic stability in line with Sustainable Development Goal 1 (SDG 1) of
eradicating poverty in all its forms everywhere. India's agricultural sector is
essential to feed its growing population, projected to reach 1.6 billion by
2050. Crop diseases, particularly wheat rusts—yellow rust, leaf rust, and stem
rust—are significant obstacles to agricultural productivity, causing substantial
yield losses and affecting the livelihoods of millions of farmers. Therefore,
developing an automated system for recognizing and classifying wheat rust
diseases is crucial for ensuring food security and economic stability. This
study aims to design an automated system for identifying and categorizing wheat
rust diseases using advanced image processing and machine learning techniques.
We collected a dataset of wheat leaf images from agricultural fields in Punjab
and Haryana and applied noise filtering and segmentation methods to enhance
image quality. Transfer learning and deep convolutional neural networks (CNNs)
were used to develop a classifier model. The ResNet50 model achieved an accuracy
of 98.5% in classifying wheat rust diseases. By addressing wheat rust diseases
effectively, this system supports SDG 1 by enhancing agricultural productivity,
improving food security, and contributing to the economic well-being of farmers. |
Keywords: |
Deep learning, wheat rust, transfer learning, convolutional neural
networks, agricultural productivity, SDG 1, poverty alleviation. |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
USING LONGEST COMMON SUBSTRING ALGORITHM TO FIND SIMILAR QURANIC VERSES |
Author: |
MAJED ABUSAFIYA |
Abstract: |
In this paper, the longest common substring algorithm is used to find similar
Quranic verses (ayahs). To show the potential of using computation in this
context, pairs of similar ayahs that were found computationally are compared
with those that are explicitly stated in a specialized scholarly book in this
field. It was found that only about one third of these pairs exactly matched
those that are computationally found. However, most of the other pairs of
similar ayahs that were mentioned in the book showed less similarity than those
that were found computationally. The main contribution of this work is showing
the value of the proposed computational solution through finding similar Quranic
ayahs beyond those that are documented in the specialized books. |
Keywords: |
String Similarity Algorithm, Quran, Quranic Verse Similarity |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
DIGITAL TWIN ENERGY MANAGEMENT SYSTEM WITH ARTIFICIAL INTELLIGENCE INTERNET OF
THING TO SMART CAMPUS |
Author: |
TANAPEAK PEXYEAN, KOBKIAT SARAUBON2, PRACHYANUN NILSOOK |
Abstract: |
The purpose of this research study is to develop a twin energy management system
with intellectual technology to connect all things to a smart campus and to
evaluate the suitability of a twin energy management system with intellectual
technology to connect all things to a smart campus. The research is divided into
two phases: Phase 1 is the development of a twin energy management system with
intelligent technology that connects all things to be able to meet the needs of
energy management within the smart campus. Theories from documents and research
related to physical energy management. It then synthesizes and links the
relationships of all the important physical elements to design the architecture.
Phase 2 is the development of a twin energy management system with intelligent
technology that connects all things to the world, which consists of 4 main
working system components, Part 1 is Physical part Synthesizing various areas
within the smart campus that must be equipped with smart sensors to process data
and control smart devices in that area to have an environment suitable for
learning in a smart campus. Part 2 is Energy Management, which is an important
process of managing data received from sensors to process various data, Cloud
Gateway, Streaming Data, Data Lake, Control Applications, Data Analytics, User
Engy Business logic, and Part 3 is Intelligence Technology, consisting of
Machine leaning Decision marking and Models algorithm that analyzes and
forecasts energy management intelligently. Part 4 is the Digital Twin part that
uses the dashboard to display in a digital form that is like a physical aspect
of communicating with the user. The results show that twin digital energy
management systems that design and develop the system can be used to manage
energy in smart campuses at the greatest scale. |
Keywords: |
Digital Twin, Energy Management System, Smart Campus |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
SENTIMENT ANALYSIS FOR TOURISM DESTINATION REVIEWS USING CROSS-INDUSTRY STANDARD
PROCESS FOR DATA-MINING |
Author: |
YERIK AFRIANTO SINGGALEN, SIH YULIANA WAHYUNINGTYAS, YOHANES EKO WIDODO, MUHAMAD
NUR AGUS DASRA, RUBEN WILLIAM SETIAWAN |
Abstract: |
This study investigates sentiment analysis methodologies in the tourism domain,
addressing the challenge of extracting meaningful insights from user-generated
content, specifically TripAdvisor reviews of S.E.A Aquarium and Gardens by the
Bay in Singapore. The research applies the CRISP-DM framework to develop and
evaluate sentiment classification models, including Naive Bayes Classifier
(NBC), k-nearest Neighbors (k-NN), Decision Trees (DT), and Support Vector
Machines (SVM). A key problem addressed is the class imbalance in the data,
which is mitigated using SMOTE, significantly enhancing model accuracy from
0.500 to 0.992. Performance differences across models are assessed through
comprehensive t-tests, revealing significant results with p-values ranging from
0.000 to 0.182. The models are evaluated using metrics such as accuracy,
precision, recall, AUC, and F-measure, with SVM demonstrating the best overall
performance. Additionally, a word frequency analysis highlights recurring themes
in tourist feedback. The study’s findings contribute to a deeper understanding
of sentiment analysis in tourism, offering valuable insights for improving
tourist experiences and guiding destination management strategies. |
Keywords: |
Sentiment Analysis, Tourism Reviews, TripAdvisor Data, VADER |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
OPTIMIZING CLOUD RESOURCE UTILIZATION THROUGH MACHINE LEARNING FORECASTING |
Author: |
ANGELICA KAYLEE WIESI, JOSHUA SAMUAL, MOHAMMED AMIN ALMAIAH, AITIZAZ ALI1,
TAYSEER ALKHDOUR, ROMEL AL-ALI, THEYAZN H.H. ALDHYANI, RAMI SHEHAB |
Abstract: |
This research introduces a resource utilization prediction tool tailored for
dynamic and seasonal workloads in cloud environments. Traditional prediction
methods often fall short in accuracy due to the constantly changing nature of
cloud resources and workloads. To address this gap, the research proposes a
machine learning-centric approach aimed at enhancing prediction accuracy,
thereby promoting sustainability, energy savings, and improved user experience
in line with SDG 7: Affordable and Clean Energy. The approach begins with data
collection and preprocessing, employing techniques such as Fourier Series and
Lag Features to capture temporal patterns. Three machine learning models—Gated
Recurrent Unit (GRU), Long Short-Term Memory (LSTM), and Random Forest—are
developed, trained, and evaluated using metrics like MAE, RMSE, and MAPE.
Hyperparameter tuning is conducted to optimize model performance and minimize
overfitting. The best-performing model, identified as the one-step GRU, is then
deployed using Streamlit and AWS EC2, with User Acceptance Testing (UAT)
ensuring it meets performance standards. This comprehensive approach
demonstrates significant improvements in prediction accuracy and resource
management, contributing to more efficient and sustainable cloud computing
practices. |
Keywords: |
Resource Utilization; Lstm; Gru; Random Forest; Machine Learning; Sustainability |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
SUSTAINABLE DEVELOPMENT OF SCIENTIFIC RESEARCH INFRASTRUCTURE FOR QUALITY AND
INNOVATIVE HIGHER EDUCATION IN THE FIELD OF INFORMATION TECHNOLOGY |
Author: |
LUCIANA TOTI, ELDA CINA |
Abstract: |
The scientific research of the twenty-first century necessitates robust
infrastructure, as well as seamless interaction between the acquired academic
knowledge and innovative technical services enabled by digital technologies.
Research Infrastructure (RI) embodies a symbiotic relationship among scientific,
social, and material dimensions, providing directions for new strategies
development. With the help of connections and collaborations with industry,
business, and international institutions, universities enhance their laboratory
capabilities, thereby advancing scientific research. This constructive
interaction forms a self-reinforcing cycle benefiting all stakeholders involved.
Drawing from collaborative experiences, with partner universities across
numerous EU-funded projects, we advocate for scientific researchers to keenly
identify contemporary needs and meet the expectations of diverse stakeholders.
Research infrastructure is the only way to optimize the identification of these
needs. This paper aims to identify pathways for attracting investments to
establish essential RI aligned with contemporary demands while assessing the
impact of RI on societal, economic, and scientific domains. This effort will
guide the rapid transition of the Albanian and Western Balkan Universities
towards future-oriented institutions. Additionally, our objective is to
delineate a strategic framework and establish qualitative and quantitative
benchmarks for sustainable RI development, providing a replicable case study for
universities in the Western Balkans. |
Keywords: |
Scientific Research, Research Infrastructure, Digital Technologies, Stakeholder
Engagement, Sustainable Development |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
DATA REDUNDANCY ELIMINATION FOR ENERGY SAVING IN WIRELESS SENSOR NETWORKS USING
CIRCLE INTERSECTION PROBLEM |
Author: |
D SATYANARAYANA, ABDULLAH SAID Al KALBANI |
Abstract: |
Wireless sensor networks play important roles in several monitoring and
surveillance applications. These networks have important constraints, such as
limited battery power in sensors. Hence, it is necessary to develop methods for
saving battery power and enhance the network lifetime. The protocol low energy
adaptive clustering hierarchy is an important clustering centered algorithm to
improve the network lifetime. The algorithm reduces energy dissipation during
inter-cluster data communication. Recently, evolutionary based energy efficient
clustering methods, such as Flower Pollination Algorithm based Clustering
method, Harmony Search based clustering algorithm, and Enhanced Flower
Pollination based Fuzzy Inference System algorithm are proposed to enhance the
lifetime of network. However, the methods do not consider data redundancy at the
intra-cluster transmissions. In this paper, an energy saving mechanism is
presented by considering the data redundancy at the sensor members within the
cluster. The circle intersection problem is modelled for wireless sensor
networks to decrease the energy expenditure during data transmission between
sensor nodes and cluster heads. In addition, performance analysis of proposed
method is presented. Simulations of the network are created, and the results are
presented. |
Keywords: |
Sensors, Wireless Sensor Networks, Energy Efficient Communication, Clustering
Algorithms, Circle Intersection Problem |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
ADAPTİVE SENTİMENT RECOGNİTİON İN CODE-MİXED TEXT USİNG UNSUPERVİSED LEARNİNG
ANCHORS |
Author: |
C.KUMARESAN, P. THANGARAJU |
Abstract: |
With the growing prevalence of multilingual digital communication, sentiment
analysis faces significant challenges when applied to code-mixed texts, where
multiple languages are used interchangeably within a single message. Existing
sentiment analysis models, primarily designed for monolingual data, often need
to capture the complexities of mixed-language texts due to their reliance on
large labeled datasets and the inability to interpret nuanced expressions across
languages. This paper presents a novel sentiment analysis framework that
addresses these issues by integrating XLM-R for robust multilingual feature
extraction, BiLSTM for capturing sequential dependencies, and CNN for extracting
localized features. The model introduces an anchor-based semi-supervised
learning approach, which effectively propagates sentiment labels from a small
set of labeled data to a larger pool of unlabeled data, significantly reducing
the dependency on manual annotations. Experimental results demonstrate that the
proposed model outperforms traditional sentiment analysis methods in accuracy
and F1-score, confirming its effectiveness in handling code-mixed text. This
research advances the field by offering a scalable solution for sentiment
analysis in multilingual settings. It highlights the increasing importance of
adapting natural language processing models to real-world, linguistically
diverse data. |
Keywords: |
Sentiment Analysis, Code-Mixed Text, Semi-Supervised Learning, Multilingual NLP,
Anchor Detection, XLM-R, BiLSTM-CNN Ensemble. |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
A SYSTEMATIC LITERATURE REVIEW ON BLOCKCHAIN FOR IMPLEMENTATION OF DATA
GOVERNANCE FRAMEWORK |
Author: |
MUHAMMAD FEIZAL, ROSMAYATI MOHEMAD, NOOR MAIZURA MOHAMAD NOOR |
Abstract: |
Data plays an increasingly pivotal role as it has the influence to shape the
decision-making process of an institution, ultimately impacting its
competitiveness and enhancing customer satisfaction. As technology continues to
progress, the volume of data accessible is expanding from various diverse data
sources, thereby leading to the possibility of data inaccuracies due to the
significant challenges in integrating, maintaining consistency, and ensuring
interoperability of the data. The accuracy of decision-making is greatly
influenced by the trust level of the data source and the quality of the data.
High data quality and a strong level of trust in data are essential in data
governance to guarantee the accuracy, consistency, security, and accessibility
of data. This, in turn, facilitates decision-making and resource management.
Currently, the majority of public understanding regarding blockchain is that
blockchain is the same as cryptocurrency. However, based on its characteristics,
blockchain has the opportunity to be used for data management in various fields.
The potential of blockchain apart from cryptocurrency in this case for data
management is not yet widely understood by most people. The purpose of this
paper is to summarize the development of blockchain role in supporting data
governance using a systematic literature review (SLR) method through Preferred
Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) and
state-of-the-art. A total of 449 papers are collected from Scopus, IEEExplorer,
Science Direct, and Google Scholar Search Engine. From these, 15 papers are
selected. The findings indicate that blockchain technology has the potential to
significantly improve data governance in a variety of areas, including data
value assessment, data standardization and transformation, data transparency,
data security, data sharing, and data storage. This can be done because
blockchain has the capability to utilize cryptography and smart contract
automatically. The key elements comprising blockchain-based data governance are
users, access and integration protocol, and data storage. |
Keywords: |
Blockchain, Data Governance, Systematic Literature Review, SLR, Trust, Data
Quality |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
THE DEVELOPMENT OF A BLOCKCHAIN-BASED SYSTEM FOR ELECTRONIC VOTING |
Author: |
EMAN-YASSER DARAGHMI, AHMED HAMOUDI |
Abstract: |
Elections and voting play a crucial role in the development of a democratic
society, enabling the public to express their views and participate in the
decision-making process. Voting methods have evolved from paper ballot systems
to e-voting systems to preserve the integrity of votes, ensuring a secure,
transparent, and verifiable process. Continuous efforts have been made to
develop a secure e-voting system that eliminates fraud attempts and provides
accurate voting results. In this paper, we propose the architecture of a
blockchain-based e-voting system called VoteChain. Developed to support the
existing voting system in the state of Palestine, VoteChain aims to provide
secure e-voting with features such as auditability, verifiability, accuracy,
privacy, flexibility, transparency, mobility, availability, convenience, data
integrity, and distribution of authority. The work introduces a smart contract
designed to meet the demands of e-voting, governing transactions, monitoring
computations, enforcing acceptable usage policies, and managing data usage after
transmission. The proposed system also adopts advanced cryptographic techniques
to enhance security. VoteChain features a web-based interface to facilitate user
interaction, providing protection against multiple or double voting to ensure
the integrity of the election. Furthermore, VoteChain is designed with a
user-friendly and easily accessible administrator interface for managing voters,
constituencies, and candidates. It ensures equal participation rights for all
voters, fostering fair and healthy competition among candidates while preserving
voter anonymity. |
Keywords: |
Blockchain, Smart Contracts, Consensus, E-voting, Privacy, Security. |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
CLUSTERING TYPES OF CAPTURE FISHERIES PRODUCTS USING THE K-MEANS CLUSTERING
ALGORITHM |
Author: |
NURDIN, BUSTAMI, RINI MEIYANTI, AMALIA FAHADA |
Abstract: |
North Aceh is one of the districts in Aceh Province which is rich in capture
fisheries resources. The North Aceh Regency Maritime and Fisheries Service
annually records a large volume of captured fisheries products, reaching tens of
thousands of tons with 75 fish divided into 3 types of fish, namely pelagic
fish, demersal fish and coral fish spread across 8 sub-districts. is a coastal
area. The problem in this research is that there is no system for clustering
capture fishery products in North Aceh Regency, so it will be difficult to
determine which types of fishery products are classified as low, medium and high
catches. This research aims to cluster capture fishery products in North Aceh
Regency using the K-Means Clustering algorithm to obtain types of capture
fishery products with low catch clusters (C1), medium catch clusters (C2) and
high catch clusters (C3). The stages carried out in this research began with the
preparation of research instruments and literature review, data collection and
analysis and application of the K-Means clustering algorithm. The results
obtained from applying the K-Means clustering algorithm are pelagic fish types
with a low catch of 86%, a medium catch of 11%, a high catch of 6%. Demersal
fish type with low catch 41%, medium catch 53%, high catch 6%. Types of coral
fish with low catches of 33%, medium catches of 50% and high catches of 17%. The
K-Means clustering algorithm can be used to cluster types of capture fisheries
products in North Aceh Regency. |
Keywords: |
K-Means clustering, Capture fisheries, North Aceh Regency, Clustering algorithm,
Data mining |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
SOCIAL COGNITIVE THEORY AND THE UNVEILING OF INTENTIONS IN ACCOUNTING PRACTICE
ADAPTATION WITHIN THE METAVERSE |
Author: |
LEE CHRISTABELLE, RINDANG WIDURI |
Abstract: |
The metaverse has the potential to revolutionize the business world, including
accounting. However, the adoption of this technology by accountants remains
limited. This study aims to bridge this gap by analyzing the factors influencing
accountants' intention to participate in the metaverse, focusing on
technological literacy, gender, age, institutional support, self-efficacy,
security, and trust. The sample for this study consisted of individuals working
as accountants or in the accounting/finance/audit/taxation sector in Indonesia
who have heard and are aware of transactions in the metaverse, with 184
respondents. Data was collected through questionnaires and analyzed using
SmartPLS 4.0.9.9. The study findings reveal that technological literacy
moderated by gender and institutional support does not significantly influence
accountant self-efficacy, and security does not significantly impact the
intention to participate in metaverse accounting. Conversely, technological
literacy moderated by age significantly affects accountant self-efficacy.
Self-efficacy and trust significantly affect the intention to participate in
metaverse accounting. While gender does not significantly impact the
relationship between technological literacy and self-efficacy, younger
accountants, who tend to have higher technological literacy, demonstrate greater
self-efficacy. The lack of specialized training programs in Indonesia limits the
impact of institutional support on self-efficacy. These factors are crucial
drivers for the intention to participate in metaverse accounting. Accountants
with higher self-efficacy and trust are more likely to embrace and adapt to the
metaverse. Security concerns did not significantly influence accountants'
willingness to participate, suggesting a potential lack of awareness regarding
security risks. To encourage broader adoption, fostering technological literacy,
especially among older accountants, increasing institutional support through
specialized training programs, and raising awareness about metaverse security
are essential. |
Keywords: |
Metaverse Accounting, Technological Literacy, Institutional Support,
Self-Efficacy, Intention |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
AI-DRIVEN ADAPTIVE CURRICULUM DEVELOPMENT: ENHANCING STUDENT LEARNING OUTCOMES
ALIGNED WITH THE THAI QUALIFICATIONS FRAMEWORK IN HIGHER EDUCATION |
Author: |
SOOKSAWADDEE NATTAWUTTISIT, PARALEE MANEERAT |
Abstract: |
This research addresses the growing demand for personalized curriculum design by
developing an AI-driven recommendation model that optimizes the alignment of
higher education curricula with individual student profiles. Traditional
"one-size-fits-all" curriculum models fail to account for diverse student needs,
often neglecting individual learning preferences, career goals, and the dynamic
demands of industry. Motivated by the need to bridge this gap, the developed
model employs advanced AI techniques, including natural language processing and
machine learning, to generate personalized course recommendations. These
recommendations are tailored to students' specific career aspirations, learning
styles, and academic interests. Utilizing a comprehensive dataset of student
profiles, course descriptions, and historical academic performance records, the
study integrates deep learning frameworks such as TensorFlow for data analysis
and processing. In alignment with Thailand's commitment to Sustainable
Development Goal 4 (SDG 4), the research promotes inclusive and equitable
quality education while fostering lifelong learning opportunities for all. The
model's evaluation demonstrated remarkable success: 80% of students reported
higher engagement, 85% showed improved academic performance with an average
grade increase of 8%, and 90% expressed enhanced satisfaction with their
educational experience. These results were supported by high-performance
metrics, with precision rates between 90.5% and 94.4%, F1-scores ranging from
91.1% to 93.0%, and recall and accuracy rates between 89.5% and 94.1% and 89.7%
to 93.5%, respectively. The algorithm effectively identified relevant courses
aligned with students' career goals, resulting in improved academic outcomes and
satisfaction. Additionally, students rated their satisfaction between 4.5/5 and
4.9/5 and demonstrated higher performance in courses tailored to their
interests. Feedback underscored those personalized recommendations enhanced the
relevance of learning, while also strengthening the connection between course
content and real-world applications. This model offers universities an
innovative tool to enhance student success and engagement through personalized
educational pathways, providing a competitive edge in the educational landscape. |
Keywords: |
Artificial Intelligence, Curriculum Design, Deep Learning, Thai SDG 4,
Life-Long-Learning |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
DATA INTEGRATION APPROACHES AND DATA CLASSIFICATION ALGORITHMS: A REVIEW |
Author: |
MOHD KAMIR YUSOF, WAN MOHD AMIR FAZAMIN WAN HAMZAH, SUHAILAN SAFEI |
Abstract: |
This article is focuses on study the different data types in data integration
and the most suitable approach for data integration to support the different
data types. Three types of data; structured data, semi-structured and
unstructured data will be used in data integration for specified purposes such
as data analysis, etc. However, the challenge in data integration is to provide
a unified view or standard view and improve the accuracy during data retrieving
process. The purpose of this paper is to review current approaches in data
integration and classification algorithms for classification of the data. Four
(4) current approaches in data integration which are federation, mediator,
mashup, extract-load-transform (ETL). Meanwhile, four (4) classification
algorithm which are support vector machine (SVM), naive bayes network, decision
trees, and neural network. Based on findings, mediator approach produced better
performance in term of response time, memory usage, and CPU usage compared to
federation, mashups, and ETL. Meanwhile, SVM algorithm indicates more accurate
for data classification compared to naive bayes network, decision trees, and
neural network. Finally, based on these findings, a new model in data
integration has been proposed to overcome the issues related with unified view
or standard view and accuracy in data classification. |
Keywords: |
Database, Data integration, Structured Data, Semi-Structured Data, Unstructured
Data, Classification Data |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
EXPLOITING THE EFFICACY OF THE DRUGS USED FOR ATTENTION-DEFICIT/HYPERACTIVITY
DISORDER (ADHD) TREATMENT IN ADULTS AND CHILDREN USING A NOVEL BIG DATA-DRIVEN
TIME-DEPENDENT FLEXIBLE DEEP RECURRENT NETWORK MODEL (TDF-DRN) |
Author: |
MR. RAMAKRISHNAN VARADHARAJAN, DR. P. ANBALAGAN, DR. SARAVANAN M.S |
Abstract: |
Attention-deficit/hyperactivity disorder (ADHD) is a prevalent
neurodevelopmental condition characterized by age-inappropriate symptoms of
inattention and hyperactivity, often persisting into adulthood for a majority of
diagnosed individuals. The primary objective of this study was to assess the
effectiveness of combination medications in treating patients with ADHD,
encompassing both children and adults. A comprehensive approach was taken,
involving the collection of 1500 records from 740 trials, including patient
responses. Rigorous data cleaning and partitioning were achieved using the Map
Reduce method to eliminate duplicates and address missing values. Moreover, the
Recursive Feature Elimination (RFE) technique was implemented to enhance machine
learning model performance by removing less significant features from the
dataset. To evaluate the efficacy of drug treatment for ADHD in both adults and
children, a proposed time-dependent flexible deep recurrent network (TDF-DRN)
technique was utilized. The preliminary findings suggested that combination
medication therapy could be a successful strategy for managing ADHD symptoms
across different age groups. Notably, the study revealed that optimal medication
combinations and dosages varied depending on the age group, with children
benefiting from a range of combination therapies and generally requiring lower
dosages than adults. Additionally, the study conducted a thorough assessment of
the safety considerations and potential adverse effects associated with
different medication combinations. |
Keywords: |
Attention Deficit/Hyperactivity Disorder (ADHD), Combinations of Drugs,
Recursive Feature Elimination (RFE), Time-Dependent Flexible Deep Recurrent
Network (TDF-DRN) |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
AN INNOVATIVE USER SIMILARITY-BASED PRIVACY PRESERVATION APPROACH |
Author: |
MARIAM RAMDI, OUMAIMA LOUZAR, OUAFAE BAIDA, ABDELOUAHID LYHYAOUI |
Abstract: |
Social networks are pivotal in various domains such as e-commerce, healthcare,
and politics, providing valuable data for numerous applications. However,
leveraging this data for tasks like recommendation systems and decision-making
often encounters challenges related to user privacy. This paper proposes a novel
approach to privacy preservation that centers on user similarity within social
network graphs. Our method addresses the dual objectives of safeguarding user
privacy and maintaining data utility. By prioritizing similarity among users,
our approach effectively reduces information loss and enhances the accuracy of
the results. This contribution is significant in advancing the responsible use
of data in social network analyses, ensuring both privacy protection and
high-quality information retrieval. |
Keywords: |
Anonymization, Similarity, Link removal, Social networks, NLP |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
Title: |
SIDELOBELEVEL MINIMIZATION BY THINNING A CONCENTRIC RING ANTENNA ARRAY USING
PARETO MULTIOBJECTIVE GENETIC ALGORITHM |
Author: |
D.D.DEVISASIKALA, D.THIRIPURASUNDARI |
Abstract: |
Recent advancements in electromagnetic problem-solving have highlighted the
efficacy of metaheuristic algorithms over traditional methods. This research
employs the Pareto Multi-Objective Genetic Algorithm (PMOGA) to optimize the
thinning of large concentric ring antenna arrays composed of isotropic elements.
The primary goal is to design a ring array that not only reduces the number of
elements but also improves performance metrics. The optimized array achieves a
reduction in the number of elements while maintaining an array efficiency (η) of
64.9% and significantly enhancing side lobe level (SLL) from -17.41 dB (for a
fully populated array with uniform excitation and spacing) to -24.84 dB. The
synthesized array pattern maintains a half power beam width comparable to that
of the fully populated array with fixed inter-element spacing. This study
demonstrates the effectiveness of PMOGA in managing the complexities of antenna
array thinning and provides substantial improvements in both efficiency and
performance, offering valuable insights for mobile and satellite communication
systems. |
Keywords: |
Sidelobelevel, Concentric Ring Antenna Array, Genetic Algorithm,Uniform
Excitation,Uniform Spacing. |
Source: |
Journal of Theoretical and Applied Information Technology
15th September 2024 -- Vol. 102. No. 17-- 2024 |
Full
Text |
|
|
|