		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 in case of review papers).

Journal of Theoretical and Applied Information Technology
June 2024 | Vol. 102 No.12

Title:	PROBABILISTIC DEEP LEARNING MODEL FOR THE LUNG CANCER DIAGNOSIS WITH FEATURE EXTRACTION AND SEGMENTATION
Author:	B SRIDHAR MURTHY, M SRIRAM, GANESAN
Abstract:	Cancer of the lung develops from cells within the lung, most commonly in the cells that line the airways (epithelial cells). Tobacco use is a major contributor to this cancer, which ranks high among the world's most deadly diseases. Exposure to environmental contaminants or a hereditary predisposition can also cause lung cancer in nonsmokers. Since early-stage lung cancer is often asymptomatic, diagnosis is often delayed until the disease has progressed significantly, leaving patients with few treatment options. The Probabilistic Fuzzy Ranking Classification (PFRC) model is presented in this paper as a new method for identifying and categorizing lung cancer from medical imaging data. The model integrates probabilistic and fuzzy ranking techniques to address the inherent complexity and uncertainty in medical images. Simulation results demonstrate the PFRC model's efficacy in accurately classifying instances within a comprehensive dataset, showcasing its robust learning capabilities. The model's configuration includes Gaussian distribution likelihoods, uniform distribution priors, and triangular membership functions for fuzzy logic parameters. With a dataset of 800 instances for training and 200 for testing, the PFRC model employs 15 extracted features for a nuanced representation of input variables. Instances of classification, feature estimation, and classification metrics such as accuracy, precision, recall, and F1 Score collectively highlight the model's strengths and areas for refinement. This research contributes to the advancement of lung cancer detection methodologies, emphasizing the PFRC model's potential as a reliable tool for improving diagnostic accuracy in medical imaging.
Keywords:	Lung cancer, Image Processing, Deep Learning, Fuzzy logic, Probabilistic Model, Ranking.
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	WHALE-OPTIMIZED PROBABILISTIC SELECTION FOR ENHANCED INTRUSION DETECTION IN CLOUD ENVIRONMENTS
Author:	RAJASHEKAR KANDAKATLA, DR.K. RAJAKUMARI, DR.M.SRIRAM
Abstract:	In todays interconnected digital landscape, ensuring robust cybersecurity measures is paramount, particularly within cloud computing environments. This paper investigates the efficacy of the Whale Optimized Probabilistic Selection (WOPS) algorithm for enhancing intrusion detection in cloud systems. Leveraging WOPS's iterative feature subset optimization, the study aims to improve classification performance metrics such as accuracy, precision, recall, and F1 score. Through comprehensive experimentation and comparative analysis with traditional approaches like Support Vector Machine (SVM) and Random Forest (RF), the study demonstrates WOPS's superiority in classification accuracy while maintaining moderate computational complexity. The findings underscore WOPS's potential as a valuable tool for bolstering cybersecurity defenses in cloud computing, offering both enhanced detection capabilities and operational efficiency. By integrating WOPS into practical intrusion detection systems, organizations can enhance their security posture and mitigate emerging cyber threats in cloud-based infrastructures.
Keywords:	Optimization, Cloud Computing, Intrusion Detection, Selection
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	DEEP LEARNING FOR UNTANGLING THE CHEMISTRY OF SCENT: A NOVEL APPROACH TO ODOUR CLASSIFICATION USING GC-MS DATA
Author:	SASEDHAREN CHINNATHAMBI, GOPINATH GANAPATHY
Abstract:	The study evaluates the efficacy of employing a deep-learning neural network, for classifying the samples from the Gas Chromatography-Mass Spectrometry(GC-MS) chromatogram dataset. By utilizing a deep learning neural network, this paper endeavours to classify chemical compositions, facilitating high-throughput compound identification and assessment. The unstructured nature of the data, variability in compound identification parameters, and the need to consider experimental conditions pose additional challenges in accurate classification. To overcome these challenges, the research implements data preprocessing techniques such as linear interpolation to bridge gaps in chromatography profiles, thereby transforming the raw dataset into a structured and informative dataset. The objectives include streamlining the integration of GC-MS data into deep learning models, improving the detection and classification of odours, and providing a framework for real-time odour recognition systems. This study delves into the intricacies of GC-MS data analysis within the realm of olfactory classification, with particular attention to the fragrances of Jasminum Sambac, Rosa Damascena, and Human Urine, encompassing both pleasant and unpleasant scents. Through exploratory data analysis, crucial variables are identified, and a novel deep-learning neural network is proposed for characterizing chemical compounds and their impact on odour classification. By pioneering the application of supervised deep learning directly on raw GC-MS datasets, this study achieves remarkable accuracy in classifying floral and human urine samples. Linear interpolation emerges as a key technique for seamless data integration and augmentation, offering valuable insights into the aromatic profiles of culturally and economically significant flowers.
Keywords:	Deep Learning, Interpolation, Neural Network, GC-MS, Feature Extraction, Feature Classification
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	A SURVEY ON COMMUNITY DETECTION: APPLICATIONS, ALGORITHMS, AND CHALLENGES
Author:	LAASSEM BRAHIM, MAMMASS MOUAD , CHERKAOUI CHIHAB-EDDINE, IDARROU ALI
Abstract:	Community detection in networks is the process of identifying groups of nodes with more connections within the group than with the rest of the network. Community detection plays an important role in the field of complex network analysis. It can be used to understand the relationships and dynamics within a network, design better recommendation systems, and create powerful network visualizations. This paper surveys the most important approaches in this field and classifies them into categories based on their fundamental working principle. We have briefly discussed real-world community detection applications. Moreover, the paper discusses the strengths and limitations of each approach and the challenges and future research directions in this area. Overall, this article provides a comprehensive overview of the current state of the art in community detection and serves as a valuable resource for researchers in this field.
Keywords:	Social Network, Complex Network, Community Detection, Graph Clustering
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	SYSTEMATIC LITERATURE REVIEW ON ETHICAL CONSIDERATION IN MULTIMEDIA PROFESSIONAL PRACTICES
Author:	FAAIZAH SHAHBODIN, CHE KU NURAINI CHE KU MOHD, NUR RAIDAH RAHIM, ULKA CHANDINI PENDIT, LAKSMI DEWI, HELMI MOHD KASIM
Abstract:	The rapid advancement in technologies provide the different method of presenting information to the audience in multimedia. However, this leads to the needs to review the current ethics challenges faced by the multimedia professional. This research aims to give insight in the ethical consideration including the challenges faced by the professionals and provide a guideline in ethic decision making by having a systematic literature review. Privacy issues, balance between realism and addiction and truthfulness of information are the ethical challenges faced by the multimedia professionals. Cooperation among the community and involvement of the audience in the development process are the guideline provided to minimize the risk of the ethics issues faced by them.
Keywords:	Multimedia, Ethics, Privacy, Immersive, ICT
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	HEALTHCARE QUESTION ANSWERING SYSTEM IN BENGALI – A PROPOSED MODEL
Author:	ARGHYADEEP SEN, DR. SATYA RANJAN DASH, DR. MANAS RANJAN PRADHAN, SHANTIPRIYA PARIDA
Abstract:	The rule-based Question Answering System was part of the manual input of rules and the development of rule-based models to implement an answering engine. However, manually formulated rules and knowledge of each language became the disadvantages of the rule-based approach, whereas current studies on low-resource languages lack expertise and suitable datasets in such languages. There are existing studies to obtain quantity and historical information from a given text; however, these studies were developed for educational purposes, and these models can respond only to simple questions. This study is focused mainly on obtaining healthcare related queries from patients’ perspectives. The proposed model of the Question Answering System would be able to answer patients’ queries from a set of prescriptions and medical reports. Moreover, this model can provide the necessary details about the patient to the medical personnel to understand the patient’s condition without having to comb through a long medical history of reports. The Question Answering System would be able to resolve the issues with doctors’ messy handwriting, recommend specific medicine and dosage to the patients, and help common patients understand their condition and provide support for their well-being.
Keywords:	Question Answering System, Low Resource, Health Care, Medical Question Answering Dataset
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	TEXT STEGANOGRAPHY BASED ON UNICODE CHARACTERS AS MARKER IN INDONESIAN EXCEL FILE
Author:	KUKUH ADI PRASETYO, ROJALI
Abstract:	There are many applications to simplify office activities in Indonesia today. When the confidential information file from the office application has been spread, employees will definitely be confused about which one of the several users of the office application has downloaded the file. The message as a mark was given to the confidential information file using steganography based on Unicode characters utilizing the Latin letters that appeared most frequently in Indonesian (a, n, e, and i). The letters that appeared most frequently were made to have a different Unicode but with the same Latin letter appearance to represent certain binary (2-bit, 3-bit, or 4-bit binary). In using steganography, the results of Security Ratio, Size Increasing Ratio, and Capacity were measured and Invisibility result was also seen. The message as a mark in the confidential information file was successfully inserted and the best steganography algorithm in this research was steganography based on Unicode character which used 4 letters to represent 3-bit binary producing 100% for Security Ratio, no Size Increasing Ratio (0% for SIR), and 1954.76 bits for Capacity (increased compared to steganography based on Unicode character which used 3 letters to represent 3-bit binary, 3 letters to represent a 2-bit binary, and 4 letters to represent 2-bit binary)
Keywords:	Steganography, Unicode, Character, Letter, Excel
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	ENHANCING COLLABORATION IN PROVIDING SCIENTIFIC WRITING LEARNING RESOURCES THROUGH THE DEVELOPMENT OF A LEARNING OBJECT REPOSITORY
Author:	SUNTORO, IDA ZULAEHA, HARI BAKTI MARDIKANTORO, TOMMI YUNIAWAN
Abstract:	This study aims to develop a Learning Object Repository (LOR) prototype for scientific writing. The research design follows the Waterfall approach which consists of four stages, namely analysis, design, coding, and testing. This study was conducted from June to December 2023, involving 136 students and 6 lecturers at Buddhist Higher Education Institutions in Indonesia. The results showed that: (1) Stakeholders showed urgency towards the development of scientific writing LOR that can be used in learning and accommodating teaching materials in the curriculum. LOR users are classified into two groups, namely knowledge seekers and knowledge holders; (2) Back-end and front-end designs are carried out to facilitate users in managing data in the form of Learning Objects (LO) in LOR. These LOs are in the form of scientific writing teaching materials in various formats, such as courses, learning modules, learning videos, e-books, learning infographics, and learning portfolios, which are presented in PDF, image, video, zip, excel, and URL formats; (3) The results of testing all LOR system functionality are declared valid, indicating that this system is ready to use and has been tested on a wider scale. Users gain positive experiences from utilizing LOR as a means of collaboration in providing learning resources for scientific writing. However, to improve the effectiveness and user acceptance of LOR, in-depth evaluation of user satisfaction, development of additional features, content development, development of a knowledge management model, exploration of the integration of the latest technology, and analysis of sustainability and maintenance are needed.
Keywords:	learning object repository, learning resources, scientific writing
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	AUTOMATED BRAIN TUMOR DETECTION WITH GLCM-BASED FEATURE EXTRACTION AND PCA FOR DIMENSION REDUCTION AND CLASSIFICATION USING MACHINE LEARNING
Author:	PAVAN KUMAR PAGADALA, D YASO OMKARI, P SREE LAKSHMI, CHRISTINE DEWI, STEPHEN APRIUS SUTRESNO, HENOCH JULI CHRISTANTO
Abstract:	The pursuit for accurate brain tumor detection and classification remains essential in medical imaging applications. Leveraging modern machine learning techniques is crucial for achieving precise and automated diagnosis. This study introduces an automated system for brain tumor detection and classification using open-source Kaggle Dataset magnetic resonance imaging (MRI) scans. The system employs Otsu thresholding for image segmentation, GLCM-based feature extraction for effective feature engineering, and PCA (Principal Component Analysis) for dimensionality reduction. Notably, the salient features extracted by GLCM undergo classification using an adaptive machine learning algorithm known as Adaptive SVM (Support Vector Machine), enhancing the classification process. With a focus on performance evaluation, the proposed algorithm, incorporating adaptive SVM as a classifier, is rigorously assessed against existing methodologies. Remarkably, the experimentation results showcase exceptional accuracy, with the proposed method achieving an impressive 98.3% accuracy in both detecting and classifying brain tumors. This superiority over previous approaches underscores the efficacy of the combined GLCM, PCA, and SVM methodology in brain tumor classification and detection, offering promising advancements in diagnosis. The findings contribute significantly to the burgeoning field of machine learning in medical imaging, emphasizing the potential of adaptive SVM as a valuable tool for enhancing diagnostic precision. Notably, the attained accuracy of 98.3% surpasses that of existing works, further cementing the proposed system's efficacy in clinical practice..
Keywords:	Principal Component Analysis (PCA), Machine Learning, Gray level Co-occurrence Matrix (GLCM), Otsu, Accuracy, Support vector Machine
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	DEEP LEARNING BASED MALWARE DETECTION
Author:	T. SUSHMA, SIRISHA NARKEDAMILLI, MADHAVA RAO CHUNDURU, VADDEMPUDI SUJATHA LAKSHMI, G. BALU NARASIMHA RAO, PRABHAKAR KANDUKURI
Abstract:	Malware detection is a critical aspect of cybersecurity, aiming to identify and mitigate malicious software designed to harm or exploit any programmable device or network. Traditional methods of malware detection, such as signature-based techniques, have limitations in dealing with the sophisticated and rapidly evolving nature of modern malware. This paper explores the application of deep learning, a subset of artificial intelligence, in enhancing malware detection capabilities. By leveraging deep learning models, which can automatically learn and extract features from data, we can improve detection accuracy and adapt to new, unseen malware. This research reviews various deep learning architectures and methodologies employed in malware detection, evaluates their effectiveness, and discusses future directions and challenges in the field.
Keywords:	Malware detection, deep learning, cybersecurity.
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	DETECTIING REDUNDANT TEST CASES USING DEEP LEARNING
Author:	RAGHAD J, DARABSEH, AHMAD A, SAIFAN
Abstract:	Software testing encompasses the examination of various data scenarios to assess output and observe software behaviour. However, comprehensive testing of all software cases poses challenges due to its intricate and complex nature. This paper is dedicated to the identification of redundant test cases through the application of deep learning techniques. Four distinct deep learning algorithms—Convolutional Neural Network (CNN), Deep Belief Network (DBN), Deep Neural Network (DNN), and Long-Term Memory (LSTM)—were employed in this study. These algorithms were applied to three datasets: Common Utils for Rapied, JSOUP, and Junit. The outcomes affirm the effectiveness of deep learning algorithms in pinpointing redundant test cases. The results demonstrated that the deep neural network (DNN) is able to detect repeated test cases, which ultimately leads to fewer test cases. Compared with other algorithms of deep learning algorithms, it was found that the deep neural network (DNN) is able to cover the test cases, and it has reached a relatively high accuracy, with a result of 82.66%
Keywords:	Test Case Reduction, Redundant Test Cases, Deep Learning, Deep Neural Network, Deep Belief Network, Convolutional Neural Network, Long-Term Memory.
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	DATA INTEGRITY CONCERNS, REQUIREMENTS, AND PROOFING IN CLOUD COMPUTING
Author:	NABEEL AL-MILLI, ZEYAD JOBAIR, MOHAMMAD RASMI AL-MOUSA, ALAA AL-SHAIKH, 5MAHMOUD ASASSFEH, RAED ALAZAIDAH, ESSAM AL-DAOUD
Abstract:	Cloud computing enables users to maintain their data on remotely a cloud server and stay safe from harmful threats, such as impersonation attacks. Users opt for storing their data in external cloud storage, which provides them with usage and storage analysis services. Cloud computing is utilized to store vast volumes of images due to the rapidly growing use of images with an increasing level of detail brought on by improvements in images. The cloud has become a reality, nevertheless, there are many security issues or concerns that surround it, such as data integrity and unauthorized access, to name a few. Many provable data possession (PDP) and proofs of retrievability (POR) approaches have been put forth to evaluate the data integrity in the cloud. Moreover, distributing the data that is kept in the cloud amongst numerous users is still a problem because the company managing authentication and authorization for the cloud services is dishonest. In contrast to other publications that focus on the cloud as a whole, this paper lists concerns associated with data kept in cloud storage and resolutions to those matters.
Keywords:	Data Integrity, Cloud Computing, Data Security, Cloud Security, Integrity Protocols
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	HIGH-DIMENSIONAL DATA-DRIVEN PNEUMONIA DIAGNOSIS USING ANFIS
Author:	VEERA SWAMY PITTALA, PRANEETH CHERAKU, PARASA SOMARAJU, RAMESH BABU PITTALA, PEDAPUDI NAGABABU, KANDIMALLA GOPI
Abstract:	The present research paper presents a transformative diagnostic framework for pneumonia employing an Adaptive Neuro-Fuzzy Inference System tailored particularly for clinical high-dimensional data. The ANFIS device combines the exclusive qualities of fuzzy logic and the flexible nature of neural networks. It can analyze complicated patient data. Our extensive evaluation across multiple clinical datasets demonstrates the unmatched accuracy rate of diagnosis, which exceeds 95 percent, an accuracy rate exceeding 90%, and a similar sound recall rate, resulting at an F1 score of 0.92. A ROC-AUC of 0.98 indicates that our model, with an excellent ability to differentiate pneumonia presentations from a healthy, nuanced situation. This is a game changer for clinical diagnostics industries and suggests the system’s implications for pneumonia detection with high accuracy. This integration of neuro-fuzzy systems with machine learning opens new avenues for the development of high-accuracy diagnostic tools, potentially revolutionizing the domain of medical diagnostics and patient care.
Keywords:	Adaptive Learning, Clinical Data Analysis, Diagnostic Accuracy, Fuzzy Logic, High-Dimensional Data, Neuro-Fuzzy Systems, Pneumonia Diagnosis, ROC-AUC.
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	CHAT-SQL: NATURAL LANGUAGE TEXT TO SQL QUERIES BASED ON DEEP LEARNING TECHNIQUES
Author:	MAJHADI Khadija, MACHKOUR Mustapha
Abstract:	The conversion of a text to a Structured Query Language (SQL) is a complex process that faces multiple challenges and a variety of problems. This is because the extraction of information stored in these databases requires the use of queries expressed in terms of a database query language, such as SQL. In addition, the Natural Language Interface to Database (NLIDB) is one of the most traditional applications of the NLP field that enables end-users to easily fetch data from databases. Recently, it has gained widespread attention, mainly because of the current success of Deep Learning techniques. The dominant NLIDB systems use the sequence-to-sequence approach. It is based on Long Short-Term Memory (LSTM) networks that include an encoder and a decoder method. In this article, we will tackle first the recent encoder/decoder approaches and analyze their pros and cons. Then, we will conduct an introductory summary of our suggested model for the NL to SQL problem. Namely, how this model can outperform the already existing solutions to enable it to manage the complex Natural Language questions-to-SQL generation queries in different contexts and cross-domain datasets. For this purpose, our work in this context will be focusing on facilitating access to the information stored in a database, by constructing a model that takes as input the natural language questions and translates them automatically to a structured language query. As a result, this model will offer a large number of database users simple and unlimited access to data with no need to learn any Database Query Language.
Keywords:	Natural Language Processing (NLP), Machine Learning (ML), Deep Learning (DL), Natural Language Interface to Database (NLIDB), Long Short-Term Memory (LSTM), Structured Query Language (SQL).
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text

Title:	MONOLITHIC HYBRID THERAPY RECOMMENDER SYSTEM FOR AUTISTIC CHILDREN USING MACHINE LEARNING
Author:	KUSUMALATHA KARRE, Dr.RAMADEVI Y
Abstract:	Everyday tasks have been revolutionized by Machine Learning (ML), which has become essential in many different sectors. Its capacity to forecast and evaluate current data has greatly increased productivity across a range of industries. In order to support the development of critical skills in autistic children, this paper presents a Monolithic hybrid Therapy recommender system that offers customized therapy recommendations. It smoothly combines content-based filtering with multi-criteria collaborative filtering. Based on the severity of the symptoms, content-based filtering recommends therapies. This information is then fed into Multi criteria collaborative filtering, which creates cohorts of related therapies using a variety of similarity metrics. Multi Criteria approach for collaborative filtering is used for the first time to find similar children in all aspects. It ensures to find similar children and most common therapies. The results of Multi criteria collaborative filtering are then sent to a priority generator, which prioritizes treatment suggestions according to symptom severity. This novel combination represents a major breakthrough in the area, by customization of therapy recommendations for children with autism outperforming traditional content-based and collaborative filtering recommender systems with a precision of 80%. When compared to conventional methods, the approach suggested saves time and effort by recommending suitable therapy for autistic children, thereby benefiting physicians, parents, and caregivers.
Keywords:	Recommender System, Content based Filtering, Multi criteria collaborative Filtering, Monolithic Hybrid Recommender System.
Source:	Journal of Theoretical and Applied Information Technology 30^th June 2024 -- Vol. 102. No. 12-- 2024
Full Text