		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

September 2023 | Vol. 101 No.17

Title:	AN EFFICIENT FILTER BASED FEATURE RANKING AND NON-LINEAR ENSEMBLE LEARNING FRAMEWORK FOR SOFTWARE RELIABILITY ESTIMATION
Author:	ANUSHA MERUGU, DR. M. CHANDRA MOHAN
Abstract:	As the size and complexity of the software systems are increasing day-by-day, the software reliability is an essential research concern for both software developers and clients. Software reliability ensures that the software products are reliable and error-free during the product deployment and software testing phases. Most of the conventional reliability estimation models are independent of heterogeneous datatype with homogeneous non-homogeneous poison process measures. In this work, an efficient reliability feature ranking based multi-class ensemble classification framework is implemented for reliability estimation process. In comparison to the traditional software reliability models, experimental results showed that the proposed model has a high software reliability prediction rate with less error and high accuracy.
Keywords:	Software Reliability, Feature Ranking, Ensemble Learning Model, Classification.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	AN EFFICIENCY OF DAWG+ ALGORITHM IN DATA FILTERING FOR DATA INTEGRATION
Author:	MIMI HARYANI TASANI, MOHD KAMIR YUSOF, WAN MOHD AMIR FAZAMIN WAN HAMZAH
Abstract:	Data integration is one of the most important components in organization especially for decision maker. The challenge in data integration is to provide a standard view based on different data format, different data schema, etc. A standard view is needed to allow different applications to access the data from different data sources efficiently. One of the issues in providing a standard view is to remove or clean special or unused characters from different data sources. Three (3) different data filtering algorithms have been used in data integration; Knuth Morris Pratt Algorithm (KMP), Boyer-Moore (BM) Algorithm and Backward Non-Deterministic DAWG Algorithm. DAWG algorithm is proven the best of data filtering in term of processing time compared to KMP and BM. In this paper, a modification of the DAWG and named it as a DAWG+ has been proposed. In this algorithm, data will be filtered and removed based on designed a library table. Modification of DAWG algorithm is needs to produce the library table. Three different datasets; SigmodRecord, NASA and DBLP will be used for experiments purposes. Based on the experiments, DAWG+ produced better result in term of data converting response time and query processing response time compared to DAWG. In conclusion, DAWG+ is proven to improve the efficiency during data retrieval process.
Keywords:	Data Integration, Data Filtering, Knuth Morris Pratt, Boyer-Moore, DAWG, Algorithms
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	PREDICTION OF THE INSURANCE CONTRACT RENEWAL FOR VEHICLE
Author:	FATMA MANLAIKHAF
Abstract:	Convincing customers to renew their insurance contract is an important task for every insurance company. To do so, prior to the expiry date, insurance companies call and send out a series of SMS to their customers in order to persuade them. However, in Morocco, we have an average of 40% nonrenewal of vehicle insurance for contracts lasting from three to six months, and these type of contracts represent 75% of the customers. In our study, we want to focus on the reasons for non-renewal in order to predict whether a customer will be able to renew their contract or not, as well as optimize sales productivity to focus on customers with a higher chance of renewal, then persuade the customer to ensure a longer duration and to provide a better proposal. Our goal in this work is to use Machine Learning (ML) in the field of vehicle insurance to extract meaningful information from data. We used approximately 4000 customer portfolios to investigate the various non-renewal issues. In this scenario, we predict contract renewal using ML methods such as logistic regression, decision tree, random forest, K-NN, and SVM. We also examine and compare the performance of various models. In terms of accuracy, kappa, and AUC, SVM beat other approaches, with values of 0.96, 0.81, and 0.90, respectively.
Keywords:	Machine Learning, Insurance Contract, Artificial Intelligence, ML Models, Classification Approach.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	NAIVE BAYES SENTIMENT ANALYSIS ON PERCEPTIONS OF HALAL CERTIFICATION: A CASE STUDY ON MIXUE INDONESIA
Author:	MULYANI KARMAGATRI, ISTON DWIJA UTAMA, HELEN YOGIE, JESTINE LEMENA CHEN, HAZMILAH HASAN
Abstract:	Indonesia, as a predominantly Muslim country, experiences a high demand for halal products due to its Muslim consumer base. This research focuses on understanding the perceptions of Indonesian society towards businesses, particularly in the food and beverage sector, that lack halal certification. The specific company under investigation is Mixue Indonesia. The study utilizes sentiment analysis through the Naive Bayes method, aided by RapidMiner software, to analyze public comments on Twitter regarding Mixue Indonesia's halal certification issue. Out of the 200 comments analyzed, the sentiment analysis revealed that approximately 122 comments expressed negative sentiments towards Mixue's halal certification. The findings indicate a notable concern and dissatisfaction among the Indonesian community regarding Mixue Indonesia's halal certification status. To address this issue, it is recommended that Mixue Indonesia takes immediate steps to obtain halal certification for their products. Additionally, future research should explore the reasons behind the negative sentiments and further investigate the impact of halal certification on consumer behavior and purchasing decisions. Understanding these factors will enable companies like Mixue Indonesia to meet the demands and expectations of their Muslim consumer base effectively.
Keywords:	Sentiment Analysis, Halal Certification, Naivebayes, Mixue Indonesia
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	COOPERATION OF TWO LOGISTIC REGRESSION MODELS FOR DECODING LINEAR BLOCK CODES
Author:	CHEMS EDDINE IDRISSI IMRANE, SAID NOUH1, BELLFKIH EL MEHDI, MOHAMMED EL ASSAD, ABDELAZIZ MARKAZ
Abstract:	Error-correcting codes (ECCs) play a vital role in protecting data against corruption in storage systems and random errors due to noise effects in communication channels. Communication protocols typically employ various techniques to detect and correct errors in transmitted messages. Syndrome-based decoding is one of the most widely used techniques for decoding algorithms. In recent years, machine learning has gained attention for building efficient decoders that demonstrate promising results, offering new avenues for error correction. In this paper, we propose improving our logistic regression-based decoder utilizing two distinct models (2LRDec). The key distinction between LRDEC and 2LRDEC is how they manage decoding tasks. While both use logistic regression models for decoding, 2LRDEC specifically employs two distinct models, which significantly enhance its performance, especially for codes with larger lengths (n). Another critical advantage of 2LRDEC is its reduced training and execution complexity. Splitting the data and using two models in parallel, allows for faster training and execution, thus increasing overall decoding efficiency. This makes 2LRDEC an effective and efficient solution for handling larger code lengths and real-time decoding scenarios.
Keywords:	Error Correcting Code; Syndrome Decoding; Machine Learning; Logistic Regression.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	THE MIRACLE OF DEEP LEARNING IN THE HOLY QURAN
Author:	SAMY S. ABU-NASER, BASEM S. ABUNASSER
Abstract:	Deep Learning, the forefront of Artificial Intelligence advancements, has gained significant global attention for its applications in various fields. While achievements in deep learning products, such as self-driving cars, chatbots, image colorization, translations, and virtual assistants, are celebrated, the connection between deep learning and the Holy Quran remains largely unexplored. This research paper aims to bridge this gap by uncovering the concept of deep learning within the Holy Quran and providing a relevant example that aligns with modern deep learning principles. The study utilizes accuracy, F1-score, recall, and precision measures to evaluate the proposed deep learning model, achieving an impressive score of 99.06% for F1-score, recall, and precision.
Keywords:	Deep Learning; Holy Quran; Artificial Intelligence; Machine Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	ENHANCING RARE DISEASE DIAGNOSIS: A WEIGHTED COSINE SIMILARITY APPROACH FOR IMPROVED K-NEAREST NEIGHBOR ALGORITHM
Author:	SOMIYA ABOKADR, AZREEN AZMAN, HAZLINA HAMDAN, NURUL AMELINA
Abstract:	Diagnosing rare diseases is challenging because they affect only a restricted group of individuals, usually identified as one out of every 2,000 people within the European Union and no more than one out of 1,250 individuals in the United States. This makes it difficult for doctors to recognize the symptoms of these diseases. This paper focuses on the challenges of diagnosing rare diseases due to their low prevalence rates and difficulties in recognizing their symptoms. Machine learning techniques often face difficulties in classifying patients with rare diseases because of their small sample sizes, leading to biased results. They proposed a weighted cosine similarity approach as a distance measure for the k-nearest neighbours algorithm instead of the conventional cosine similarity to address this issue. The use of genetic optimization to select the best weights for the weighted cosine similarity. The Rare Metabolic Diseases Database was used as a case study, and the results demonstrated that reducing the classification bias between majority and minority classes improves all classification performance measures. However, as the number of classes and imbalance ratio increase, the approach's effectiveness decreases, eventually reaching zero. Future work will focus on reformulating the g-mean to smooth its values and avoid assigning a zero score when all class instances are misclassified.
Keywords:	K-Nearest Neighbor, Cosine Similarity, Imbalance Data, Genetic Algorithm, Imbalance Ratio
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	A MULTIFRACTAL APPROACH FOR TEXTURE CLASSIFICATION APPLIED TO THE KIDNEY ULTRASOUND IMAGES
Author:	TAHIRI ALAOUI M, KORCHIYNE R,FARSSI S M, TOUZANI A
Abstract:	The realization of diagnostic decision support systems in medical imaging, in particular ultrasound imaging, is a field that continues to experience more and more success thanks to the development of image processing tools allowing segmentation and texture characterization which lead to pattern recognition with a minimum of error. Indeed, concerning the characterization of the texture of ultrasound images, in addition to statistical techniques, several methods based on new approaches are proposed. Indeed, multifractal analysis in texture characterization is becoming increasingly efficient due to multifractal modelling and powerful methods in estimation of irregularities functions and multifractal spectrum. But of all methods proposed for texture characterization, in particular methods using multifractal tools, the extraction of discriminating texture features remains a major challenge. In this respect, as a first main contribution, we propose a multifractal approach, based on two multifractal descriptions, for texture characterization that we used, in the second main contribution, to characterize textured ultrasound images of the kidney. The first texture multifractal approach is based on local information that is the singularity. Indeed, the originality of our contribution here is to use pixel singularity to build the singularity-level matrix and the binarized image(homogeneity/discontinuity), from which we extract new texture features. While the second one is based on global information provided by the multifractal spectrum. In fact, the originality of our contribution here is to use the Hausdorff multifractal spectrum that gives a better estimation of the multifractal spectrum. In this regard, to make the analysis of the spectrum possible, we had to preprocess it through smoothing, enabling us to keep the most important information of the spectrum from which we extract new texture characteristics by analyzing the shape and position of the smoothed multifractal spectrum of Hausdorff. Finally, as a second main contribution, we will evaluate the potential of our proposed multifractal features to characterize textured ultrasound images of the kidney. Having more reproducibility of the texture features first require a good choice of Choquet capacity to calculate the irregularities but also selecting a more representative region of interest (ROI) to analyze by carrying out an adapted virtual puncture in the kidney representative components. We proceed to exclusive evaluation of the four proposed methods before evaluating the combined method of the four proposed methods and the combined one. The results of the supervised classification, using three classes of images (young, healthy, glomerulonephritis), are interesting and promising since the classification accuracy reaches about 84%. This encourages conduct further research to yield better results.
Keywords:	Multifractal Analysis, singularity, Smoothing, Hausdorff Spectrum, Texture Analysis, Kidney Image, Classification.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	THE PROFITABILITY OF INVESTING IN FINE ART: AN ANALYSIS OF RESALE DATA FROM SOTHEBYS, CHRISTIES, AND PHILLIPS
Author:	ANNA VASINA, VALERY BUKANOV, VALERIA KOLYCHEVA, ALEXANDER SEMENOV, DMITRY GRIGORIEV
Abstract:	The main goal of the presented study was to determine the profitability of investing in unconventional assets, namely, fine art paintings. To do this, we have gathered and analyzed data on the resales of paintings from Sotheby's, Christie's, and Phillips from 2003 to 2021. We calculated the annual effective rate of return for each painting and divided the data into profitable and unprofitable investments. We also analyzed the impact of various factors, such as the initial sale price, the length of ownership, and the annual effective rate of return, on the resale price of the paintings. We used correlation analysis to debunk some myths about investing in art, namely myths about shelter-investments and the masterpiece effect. We also compared the annual effective rate of return for paintings with that of traditional financial assets like government bonds and gold. The results showed that investing in paintings can be quite profitable and is a good investment strategy.
Keywords:
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	BARRIERS TO CLOUD COMPUTING ADOPTION AMONG SMEs IN THE MIDDLE EAST: A SYSTEMATIC REVIEW
Author:	ZAKARIA ALRABABAH
Abstract:	The advancement of technology is significant in shaping business development by providing different means of doing business. The geometric increase in intense market rivalry and a fast-evolving business environment have caused businesses across all industries and sectors to employ information and communication technologies (ICTs) to improve their business operations and increase their business value. This paper examines the barriers to adopting cloud computing by small and medium-sized enterprises in middle eastern countries. The study used the PRISMA protocol to document the method of analysis, inclusion, and exclusion criteria. The study also used three relevant databases (Emerald, Google scholar, and Sustainability) to identify relevant journal articles which used Qualitative, Quantitative, and/or mixed research methods. The included articles must have been published in English between 2016 and the first quarter of 2023. More importantly, studies that did not focus on SMEs in the middle east were excluded from the review. After thorough literature review, it was found among others that there needs to be more awareness of the importance of industry 4.0, financial constraints, and lack of infrastructure. Recommended panaceas to the identified barriers were also presented.
Keywords:	Cloud Computing, Smes, Middle East, Barriers, Adoption
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	MAXIMIZING THE USE OF BLOCK PAYLOAD IN BLOCKCHAIN-BASED E-VOTING
Author:	ANDI, GEDE PUTRA KUSUMA
Abstract:	Nowadays, internet users are increasing in Indonesia so that it is a good opportunity if traditional paper-based voting is replaced with online voting called e-voting. E-voting has many advantages including environmentally friendly and increased efficiency. However, security issues have always been a barrier to its implementation because e-voting uses a centralized system. Therefore, many researchers propose a combination of e-voting with a decentralized system called blockchain. They believe that voting data becomes more secure because of the immutability of blockchain. However, the process of storing and validating on blockchain is quite slow, so it is not yet feasible to combine it with e-voting. Hence, there are researchers who propose validation from a centralized system such as a fingerprint, but this validation is not secure enough because it can be tampered with. There are also researchers who propose reducing or enlarging the block size on the blockchain to speed up processes on the blockchain. However, increasing the block size slows down the propagation of the block to other nodes. On the other hand, reducing the block size escalates the block composition time to clear all the transactions from the memory pool called mempool. Eventually, e-voting remains inapplicable although e-voting is combined with a modified blockchain in particular blocksize. Thus, we propose optimizing the capacity utilization of a block without changing the capacity of the block itself. Experimental results reveal that the more transactions that can be contained in a block, the faster the data search process, especially in the validation process of uniqueness feature in e-voting and vice versa.
Keywords:	Voting, E-voting System, Blockchain System, Block Payload, Uniqueness Validation
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	IMAGE FEATURE ENCODING USING LOWNERIZATION TENSOR
Author:	K. PAVAN KUMAR, Y. SURESH, PESN KRISHNA PRASAD, CH V K N S N MOORTHY, A. S. SAI PRASAD
Abstract:	In recent days, technology plays a major role in real-time applications of current usage and at the same time security is one of the critical task in order to utilize and access data as well applications/tools. In this scenario, vulnerability is the most important factor for any real-time application. There is a need to minimize the vulnerability of any application up to date. In majority applications, traditional human traits like Face and fingerprint can be treated as one of the prime data to provide security, through these patterns with conventional methods of access is major risk to deal. A multimodal authentication method has been proposed by these two human traits for authentication by applying a two level of process for the analysis, encoding and Decoding, 1) to preprocess the images and then extract the features using Non-Negative Matrix Factorization (NNMF) and 2) to encode and decode the extracted features using Tensor based Lownerization method. The major contribution of this method is to minimize the vulnerability of traits/patterns which extracted the features by applying NNMF method. The extent of the proposed method has been validated with the results that are obtained in two ways. In one way, after extracting the features fusion can be applied. In another way fusion is not applied. Here the two distinct metrics viz., Euclidean Distance and Mean Square Error are used. When compared to the existing papers the proposed encoding and decoding method gives better security. Mean Square Error Distance gives better results when comparing to the Euclidean Distance.
Keywords:	Multimodal Authentication, Preprocessing, Tensors, Lownerization Tensor, IF Encoding
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	SABA: SECURE APPROACH BASED ON ANOMALY AND SIGNATURE-BASED DETECTION MECHANISM FOR DETECTING ABNORMAL ACTIVITIES IN BLOCKCHAIN NETWORK
Author:	HALA A. ALBAROODI, MOHAMMED ANBAR
Abstract:	In recent years, blockchain technology has undoubtedly experienced broad use. Apart from its initial usage in cryptocurrency, it is now employed in healthcare, real estate, smart contacts, and other fields. However, many blockchain security vulnerabilities have been caused by the incorrect implementation of the technology. As a result, the Blockchain may become insecure, allowing attackers to carry out a variety of Blockchain-based attacks. Suspicious behaviour is expected may exist because of the presence of blockchain attacks. Therefore, detecting suspicious behaviour may detect different types of Blockchain-based attacks. Thus, this paper aims to propose the Signature and Anomaly approach (SABA) to detect suspicious behaviour in a Blockchain environment based on Indicators of Compromise (IOCs). SABA consists of components as follows: the first layer is the Blockchain application (threats detector; and APIs), the second layer is the protocol layer (decentralized protocol), and the third layer is the data layer or can call it to overlay network (SBAB fork module; SBAB transactions filter; and SBAB threat database). The proposed approach is detailed in-depth and proven by experimental and analytical findings showing the quality and practicality of Blockchain Signature Based and Anomaly Based detection techniques.
Keywords:	Blockchain; Security; Authentication; Cloud Computing; Indicators of Compromise; Intrusion Detection Systems; Signature Detection Based; Anomaly Detection Based.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	NER-NET: A NEW DEEP LEARNING ARCHITECTURE WITH HYBRID OPTIMIZATION FOR NAMED ENTITY RECOGNITION
Author:	P SHEELA GOWR, N KUMAR
Abstract:	Identification of textual and visual information in a document plays a pivotal role in comprehending the information contained in it, thereby providing an effective way to analyze the document. Information Extraction (IE) is an attractive research area that mainly targets developing techniques for analyzing rich documents. Named Entity Recognition (NER) is the prime process in IE, and it is a sequence-labeling process, wherein unstructured data is considered and the data is mapped to pre-defined labels. In this work, a novel deep learning architecture is developed for performing NER, and it is aimed at identifying the numerical entities in the input text. Here, a novel deep learning network named NER-Net is proposed which includes various layers, like tokenization, Inverse Document Frequency (IDF), Convolutional Neural Network (CNN), Bidirectional Long Short Term Memory (BiLSTM), attention, and NER layers. Further, a Fire Hawk African Vultures Optimization Algorithm (FHAVOA) is proposed for optimally tuning the layer dimension of the NER-Net. Moreover, the evaluation of NER-Net-FHAVOA shows that accuracy of 0.936, Mean Square Error (MSE) of 0.087, Root MSE (RMSE) of 0.294, ‎Positive Predictive Value (PPV) of 0.909, and ‎Negative Predictive Value (NPV) of 0.877 are attained, thus revealing its superiority.
Keywords:	Fire Hawk Optimization, African Vultures Optimization Algorithm, NER-Net, Convolutional Neural Network, Bidirectional Long Short Term Memory.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	TRUST EVALUATION MODEL FOR SOCIAL INTERNET OF THINGS USING RESILIENT APPROACH
Author:	CHETHAN RAJ C, DR.J HANUMANTHAPPA, SHARATH KUMAR G N, INCHARA G P
Abstract:	Social Internet of Things is a trend in the technology which allows to the add objects to the network through which communication is possible using unique object relationship and ability to transfer the data in a network. Internet of Things is able to achieve more efficiency in decision making, Social internet of things is a subset of Internet of Things that establishes the relationship with other objects for effective communication and can improve the scalability, trust, resource management using social trust computing. Many existing models are not dynamic in nature in proving the trust with objects and user interaction and decision making process is not identifiable, the proposed Resilient Based Social Internet of Things model increases performance of evaluation with various attributes like information gain, resilience of the system, cooperativeness and trustworthiness. In SIoT trustworthiness is very important in defining reliability in user communications and interactions. The proposed experiments shows the significant improvement in the trust model for the AppClassNet data set and social internet of things data set in order to segregate trust and untrusts effectively in the network model with 92% information gain and high resilience by comparing with existing model.
Keywords:	Trust Model, Rbsiot Approach, Cluster Coefficient, Centrality, Information Gain, Cooperativeness, Betweenness
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	PREDICTION OF PNEUMONIA DISEASE FROM X-RAY IMAGES USING A MODIFIED RESNET152V2 DEEP LEARNING MODEL
Author:	D.VETRITHANGAM, PRIYANKA PRAKASH SATVE, JAMBI RATNA RAJA KUMAR, P. ANITHA, S.VIDHYA, ASHWINI KUMAR SAINI
Abstract:	The lungs play a crucial role as the primary components of the human respiratory system, making them susceptible to inflammation and impact lesions in our daily lives. Among all infections, pneumonia holds the distinction of being the most widespread worldwide, with the lungs serving as the gateway for its spread throughout the body. In hospital settings, chest X-rays emerge as the most prevalent diagnostic tool employed to accurately identify pneumonia. Physicians heavily rely on these X-ray images to make precise diagnoses and monitor the progress of pneumonia treatment. Moreover, this type of chest X-ray facilitates the detection of other conditions like emphysema, lung cancer, the positioning of lines and tubes, and tuberculosis. The challenges faced by the existing deep learning models for pneumonia prediction include high computational complexity, prolonged model training times, and a lack of efficient preprocessing techniques. These issues contribute to misdiagnosis and inaccurate predictions of pneumonia. Moreover, the lack of interpretability in many of these models further hinders their acceptance and understanding in clinical applications. This research aims to tackle the challenges presented by current techniques by proposing a customized ResNet152v2 deep learning model. The primary objective is to design and deploy this modified ResNet152v2 model for pneumonia prediction from chest X-rays, achieving high accuracy while minimizing computational complexity and reducing computation time. This model outperformed well when compared with the existing methods and produced accuracy of 99.77%, Sensitivity of 99.86%, specificity of 95.4%, and precision of 99.86%
Keywords:	Pneumonia, Resnet152v2, X-Ray, Deep Learning, Prediction.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	AN IMPROVED VLSI ARCHITECTURE FOR CFA INTERPOLATION USING CARRY SKIP ADDER
Author:	CHATLA RAJA RAO, DR. SOUMITRA KUMAR MANDAL
Abstract:	A wider range of digital devices, including as 4G/5G smart phones, digital cameras, digital notebooks, and consumer electrical items, will be able to function properly thanks to this Application of Color filter array. As a result, a linear deviation compensation approach that boosts correlation between interpolated and neighboring pixels is recommended to be used to this color filter array in an effort to enhance the performance of the reconstructed images with perfection. By prioritizing green in the color interpolation process and using a hardware-sharing methodology, we may enhance the image's resolution on both sides. Therefore, larger space in arithmetic operations and higher gate counts on VLSI architecture will be needed for the hardware sharing approach of red, green, and blue interpolation. In order to cut down on space, time, and energy requirements, the project would include a color demosaicking method that makes use of a carry skip adder as opposed to a standard ripple push adder into all current hardware sharing techniques. In this research, experiments are conducted using VHDL programming language and the synthesize capabilities of the Xilinx FPGA XC6SLX150-2CSG484 at a 200 MHz operating clock frequency to create a color demosaicking approach using a 256x256 pixel images.
Keywords:	CSKA (Carry Skip Adder), CFA (Color Filter Array), Boundary detection, Boundary Mirror Machine, VLSI.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	INTELLIGENT SECURITY SYSTEM BASED ON BIOMETRIC FACE RECOGNITION
Author:	IDA MULYADI, MUHAMMAD FAISAL, KOMANG ARYASA, SAMSURIAH, MUSDALIFA THAMRIN, DARNIATI, MUHAMMAD ADNAN
Abstract:	Biometric face recognition is used in various applications, including security systems, access control systems, time and attendance systems, and public safety applications. The application system also requires a non-physical system that can be applied to certain parts to support creating a complete security system. As an essential layer of the security system in the mixing process, authentication techniques for mixing participants may be required during mixing. Currently, the application of facial recognition is still being developed through the development of methods to increase the recognition rate of facial recognition based on facial position and expression. Based on the comparison of the algorithms used, it is known that the CAMSHIFT algorithm has the best accuracy value, with an accuracy of 99.51%. Based on the experimental results, information was obtained that the F-measure value for FASTER R-CNN and CAMSHIFT was 0.96. Therefore it can be concluded that the Faster R-CNN method is a method that can be used to detect large numbers of objects. At the same time, the CAMSHIFT algorithm is ideal for use to support the authentication process on the face. In the future, a biometric-based security system can be implemented using the extended method to get better accuracy, resulting in speed and accuracy of detection and use in more dynamic conditions.
Keywords:	Faster-RCNN, CAMSHIFT, SURF, HAAR, Face Recognition, Confusion Matrix
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	INCORPORATING ENSEMBLE AND TRANSFER LEARNING FOR AN END-TO-END AUTO-COLORIZED IMAGE DETECTION MODEL
Author:	AHMED SAMIR RAGAB, DR. SHEREEN ALY TAIE, DR. HOWIDA YOUSSRY ABDELNABY
Abstract:	Image colorization is the process of colorizing grayscale images or recoloring an already-color image. This image manipulation can be used for grayscale satellite, medical and historical images making them more expressive. With the help of the increasing computation power of deep learning techniques, the colorization algorithm’s results are becoming more realistic in such a way that human eyes cannot differentiate between natural and colorized images. However, this poses a potential security concern, as forged or illegally manipulated images can be used illegally. There is a growing need for effective detection methods to distinguish between natural color and computer-colorized images. This paper presents a novel approach that combines the advantages of transfer and ensemble learning approaches to help reduce training time and resource requirements while proposing a model to classify natural color and computer-colorized images. The proposed model uses pre-trained branches VGG16 and Resnet50, along with Mobile Net v2 or Efficientnet feature vectors. The proposed model showed promising results, with accuracy ranging from 94.55% to 99.13% and very low Half Total Error Rate values. The proposed model outperformed existing state-of-the-art models regarding classification performance and generalization capabilities.
Keywords:	Image Colorization, Ensemble Learning, Transfer Learning, Image Forensics, Colorization Detection.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	K-MEANS IN CLUSTERING THE SATISFACTION LEVEL OF CIKUNDUL HOT WATER
Author:	DUDIH GUSTIAN, MUHAMAD MUSLIH, DESHINTA ARROVA DEWI, VENTI SITI, RINA RUSTIANA, PURNAMAWATI
Abstract:	Sukabumi is a small administrative city in West Java with a land area of 4,800.231 Ha. This city, on the other hand, serves as the economic hub of the Sukabumi Regency, a much larger region. This city is home to a number of well-known tourist attractions that generate revenue for the region through both artificial and natural tourism. Cukundul Hot Spring, with its hot water acting as a magnet, is one of the natural tourist attractions that benefit Sukabumi City's local administration. However, because of the effects of PPKM implementation, the number of visitors to this tour has significantly decreased since the Covid 19 pandemic. As a result, the regional authority of Sukabumi City has seen a decrease in the amount of money it receives from these tourist attractions. The purpose of this paper is to examine the current situation and suggest some improvements. K-Means Clustering can provide a general overview of the parameter mapping investigated in this study based on visitor sample data from approximately 35 visitors. Although this study is not robust enough to serve as a benchmark due to the small sample size, it can serve as an example for managers to improve visitor services and for local governments to better manage budgets for maintenance and development, ensuring that this tour remains a favorite of tourists from within and outside Sukabumi.
Keywords:	West Java, Sukabumi City, K-Means Clustering, Visitor
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	FEATURE CLASSIFICATION BASED ON HETEROGENOUS DATA USING HYBRID MACHINE LEARNING: A REVIEW
Author:	AGUS NURSIKUWAGUS, HERI PURWANTO, DESHINTA ARROVA DEWI
Abstract:	Heterogeneous data is a dataset with various types including data type and data source. Classification of heterogeneous data is still becoming a discussion in research in the field of intelligence artificial especially in learning classification. Based on data and machine development classification, then study this still relevant done. Machine classification that is still trending now is a hybrid engine known as the collaboration technique such as a fuzzy technique and neural network. The aim of this review paper is to find opportunity research on hybrid machine learning that perform classification on heterogeneous data with multi-class targets. There are several challenges on heterogeneous data such as 1)determining algorithm normalization and text processing as Step beginning from the input layer, 2) function formation variable linguistics for every case allow existence opportunity study for linguistic processes, 3) A membership function algorithm that can adapt of the dataset used can as opportunity research, 4) finding method shaper fuzzy rule as machine inference from a neural network, 5) Process structure of every task, 6) performance like efficiency memory for processing ( management memory ), complexity (process time), and validation architecture (accuracy, precision, recall, f-measure, specification, true prediction, false prediction). The result of the research obtained is the existence opportunity for improving or developing a hybrid classification machine that can handle heterogeneous data with multi-class targets.
Keywords:	Feature, Classification, Heterogenous Data, Fuzzy, Neural Network
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	IMPLEMENTATION OF ACADEMIC INFORMATION SYSTEM USING QUALITATIVE METHOD WITH QUANTITATIVE STRATEGIC PLANNING MATRIX APPROACH
Author:	AWAN SETIAWAN, IMAN SUDIRMAN, NURMAN HELMI, DESHINTA ARROVA DEWI, ERWIN YULIANTO
Abstract:	Previous studies reveal that poorly managed Academic Information Systems can carry the risk of having a negative impact on institutions or organizations. Hence, it is important to evaluate the performance of an academic information system to measure the effectiveness and productivity of the system's success in achieving its goals. The study in this research involves a private college in Bandung city that has successfully implemented its Academic Information System for more than five years. We managed to collect 4286 SPACE (Smart Platform for Academic Environment) users and divided them into internal and external informants. The internal informants include the AACSB (Association to Advance Collegiate Schools of Business) team, QA Director, Head of Program, IT Manager, Lecturer, Secretary of Program, HCD Manager, and Users. External informant involves experts including WDA (Vice Dean for Academic), Policy Makers, Information Security Practitioners, and Academics. For the analysis tool, we use QSPM (Quantitative Strategic Planning Matrix) a well know system frequently used in the decision stage. To determine which strategy is the most effective, QSPM uses input from the IFE (Internal Factor Evaluation) and EFE (External Factor Evaluation) matrices at the input stage and the IE (Internal External) and SWOT (Strength, Weakness, Opportunities, and Strengths) matrices at the matching stage. Our findings indicate that Higher Education Institutions should take four steps to address issues with the application of system quality, information, services, and information security. First, find or hire the best IT candidate. Next, regularly review each system's features. third, make system features simpler. and finally, double-check data before system integration.
Keywords:	Effective Strategy, Academic Information System, Qualitative Method, Quantitative Strategic Planning Matrix
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	A PRACTICAL MODEL FOR USING CLASSIFICATION ALGORITHMS TO ENHANCE CARDIAC DIAGNOSIS
Author:	LIQAA ALWAN ADHOOB , RAZIEH ASGARNEZHAD
Abstract:	According to historical records, one of the most prevalent ailments is heart disease. All age groups are affected by this illness, including teenagers, adults, and the elderly. Since there is never a powerful and effective treatment that may significantly lessen the severity of this condition and there is always a failure in clinical cardiac situations, it is thought to be incurable. Individuals are at grave risk from heart disease since it has lately emerged as a severe condition that threatens people. All age groups, from young to the elderly, are often affected. The preparation of data and finding a fix for record failures are the major challenges in this area. Clinical heart failure data, where a successful, high-performance strategy was suggested to treat loss and ameliorate heart disease. The authors of this paper used the method for cleaning data and used the base classification techniques. For cleaning data, the worthwhile preprocessing step is employed to recognize the missing values and outliers in conjunction with KNN. Two experiments were handled to replace these challenges with RF, DT, and KNN. We achieved success using classification algorithms to greatly forecast and enhance the performance of heart disease. Through the outcomes of this study, this model showed a clear advantage over its competitors. The highest results obtained were 98% for all evaluation metrics. It means that we have a 3% improvement. It demonstrates that this model will make accurate predictions and enhance the performance of the data we have focused on.
Keywords:	Data Mining, Pre-processing, Heart Disease, classification, Machine Learning Techniques
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	BLOCKCHAIN UTILIZATION IN ACTIONS TO EMPOWER DIGITALIZATION OF ACCOUNTING INFORMATION SYSTEMS FOR FOR SMALL AND MEDIUM-SIZED ENTITIES IN INDONESIA
Author:	IRIYADI, MEIRYANI, THERESIA AVILLA NI PUTU SEKAR TIARA, AGUNG PURNOMO, GAZALI SALIM
Abstract:	Implementing a digital accounting information system supports efficient financial performance management for MSMEs in Indonesia. However, some MSMEs still need to apply manual financial recording and bookkeeping due to limited ability and insight regarding the digitalization of financial management. This research aims to identify and analyze the role and solution steps in realizing the empowerment of digitizing accounting information systems through blockchain to realize effectiveness for MSMEs in managing and controlling financial performance. This study uses a qualitative descriptive method based on primary data through questionnaires and secondary data from previous research obtained by studying the literature on relevant journals, articles, and documents. The results of this study indicate that there are still MSME actors who still need to keep records or bookkeeping and MSME actors who apply manual methods in managing their finances. On the other hand, the combination of digitizing accounting information systems with blockchain can optimize the financial performance of MSMEs so that comprehensive socialization is needed to create awareness and initiatives to transform towards digital. This is the first study conduct on the utilization of blockchain in actions to empower digitalization of accounting information systems for MSMEs in Indonesia.
Keywords:	Digitization, Accounting Information System, Blockchain, MSMEs, Decentralization, Technology Transformation
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	AN OCR FOR ARABIC CHARACTERS WITH ADVANCED BASELINE SEGMENTATION AND ENHANCED CONVOLUTIONAL NEURAL NETWORK FOR CLASSIFICATION
Author:	ASHIQ V M , DR. E. J. THOMSON FREDRIK
Abstract:	Human-computer communication is referred to as "Natural Language Processing (NLP)". There are several "Data Mining" along with "Machine Learning" techniques available, and the NLP goal is to allow machines to produce meaningful knowledge from natural language input. The "Arabic Optical Character Recognition (AOCR)" is a continuing process due to the obvious complex pattern and syntactic of Arabic words. Scanned documents and printed text recognition have been a major focus of AOCR research in the last several years. However, the outcomes of the AOCR study are unsatisfactory, and more work has to be done in this area. The segmentation process of AOCR is crucial for the recognition of specific Arabic texts. The character element would have a distinct representation if the fundamental form of an Arabic character is incorrectly segmented. For the segmentation of Arabic letters, we propose a "Baseline Segmentation (BS)" in this research. Every link between consecutive Arabic letters may be found along the "Base-line", which is an Arabic word's "Medium-line". Though an attempt to tackle the issues of recognizing digital Arabic Characters, including solitary numbers, characters, and vocabulary, in this research we develop a new model leveraging the "Convolutional Neural Network (CNN)" after the segmentation procedure has been completed. In particular, we enhance the conventional CNN (ECNN) model by using "Batch Normalization" and "Dropout Regularization" parameters to retrieve features that are optimum contextually. Preventing overfitting while also improving generalization is the goal of this approach. The suggested ECNN framework is designed using a multitude of convolutional layers. Multiple evaluation criteria, such as "Accuracy", "Precision", and "Recall", are used to thoroughly analyze the developed ECNN model. Further, we conduct experiments on a dataset obtained and compare the results to those obtained using preexisting EKNN and FKNN methods. The proposed ECNN approach achieves more accuracy than either the EKNN or FKNN approaches.
Keywords:	NLP, AOCR, Baseline Segmentation, ECNN, EKNN, FKNN
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	ANALYSIS OF THE MULTIHOP ROUTING ALGORITHM FOR WIRELESS SENSOR NETWORK BASED ON BAUD RATE AND PATH LOSS MODELS
Author:	G V SOWMYA, R APARNA
Abstract:	The Information gathered by the sensor node in Wireless Sensor Networks (WSNs), is sent to the sink either by multi hop transmission or by direct transmission. In multi hop transmission energy is saved as it sends the information to the nearest node and then the node hops to another node till it reaches the sink node. Selecting the optimal nearest node for sending the information is the greatest challenge in WSN. Therefore, in the proposed work, a routing protocol has been presented that selects the optimal nearest node utilizing Shannon Channel Capacity ‘C’ and a propagation model. Accurate modeling of propagation and path loss with respect to different terrains is another challenge in WSN system design and analysis. Among many metrics used to measure the performance of WSN, the proposed method uses the parameter, throughput and network lifetime. The experimental results present the type of propagation model and Shannon Channel Capacity ‘C’ that should be employed based on terrain for designing the algorithm for WSN.
Keywords:	Wireless Sensor Network, Path loss model, Shannon Channel Capacity, Routing algorithm, Multi hop, Two ray ground model, Okumura-hata model, Costa 231-hata model.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	AN EFFICIENT INTEGRITY BASED MULTI-USER CLOUD ACCESS CONTROL FRAMEWORK FOR HETEROGENEOUS CLOUD DATASETS
Author:	MOHD ANWAR ALI, NAGESH VADAPARTHI, L SUMALATHA
Abstract:	The greater part of the customary cloud based applications are shaky and hard to figure the information honesty with variable hash size on heterogeneous datasets. Due to structured data and computational memory, cloud storage systems are also independent of integrity computational and data security. As the size of the cloud information records are expanding in the general population and confidential cloud servers, it is hard to encode and translate the huge information because of document configuration and restricted respectability key size. The computational time and extra room of the ordinary quality based encryption and unscrambling models are high during the information honesty check and restricted information size. For strong data encryption and decryption, a hybrid variable-sized data integrity algorithm is implemented on heterogeneous cloud data files in this paper. For improved cloud data security, this work proposes an optimized attribute-based encryption and decryption procedure for large data files. On cloud heterogeneous data types, proposed framework outperforms conventional cloud security frameworks in terms of optimization, as demonstrated by the results of the experiments.
Keywords:	Cloud Data Security, Integrity, Encryption, Attribute Based Encryption.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	AN ANALYSIS ON ADVANCES IN LUNG CANCER DIAGNOSIS WITH MEDICAL IMAGING AND DEEP LEARNING TECHNIQUES: CHALLENGES AND OPPORTUNITIES
Author:	VAHIDUDDIN SHARIFF, CHIRANJEEVI PARITALA, KRISHNA MOHAN ANKALA
Abstract:	This abstract provides an overview of numerous studies on the identification and diagnosis of lung cancer using medical imaging and deep learning. However, techniques like the VGG16 (Visual geometry group) model, SSD (Social Ski-Driver), SVM (Support Vector Machines), CNN (Convolutional Neural Network), and CSA (crow search algorithm) have demonstrated encouraging results in accurately identifying and classifying lung cancer from CT images. Accuracy still needs to be improved. Pre-processing operations including edge identification, picture resampling, and segmentation are crucial for improving the visual appeal of input photos. The use of computer-aided diagnosis (CAD) systems can significantly increase the efficacy of diagnostic classifiers and pre-processing methods. For the identification of lung cancer, the use of deep learning models and meta-heuristic-based optimisation techniques can produce accurate and durable diagnosis models. Effective lung cancer treatment depends on early detection. The limits of the current histology-based diagnostic methods are also covered in the paper, as well as the potential of molecular biomarkers for classification. In categorising the morphology of lung cancer, selecting the optimal course of action, and forecasting the outcomes of systemic therapy for non-small cell lung cancer, machine learning algorithms have demonstrated success. A larger dataset is needed for the veracity of these conclusions, but the paper claims that doctors are increasingly employing machine learning to better understand their patients and create tailored treatment regimens.
Keywords:	Computer-Aided Diagnosis, Support Vector Machines, Convolutional Neural Network, Visual Geometry Group, Social Ski-Driver, Crow Search Algorithm.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	ARTIFICIAL NEURAL NETWORK WITH SPIDER MONKEY OPTIMIZATION ALGORITHM FOR CARDIOVASCULAR DISEASE PREDICTION
Author:	K. L. ANUSHA, JANARDHAN G, MAHANKALI SARITHA, MANNE ARCHANA
Abstract:	Globally, among the leading causes of death is coronary heart disease. Heart disease cases are rising quickly every day, thus early disease prediction is both important and dangerous. It is important to complete this diagnosis accurately and quickly because it is a challenging task. Specifically to improve diagnosis efficiency and accuracy, Machine Learning (ML) is increasingly popularity in the healthcare sector. By analyzing vast amounts of healthcare data, ML can predict diseases. Providing people and health professionals with the information they need to integrated content choices regarding preventing disease. However, the most difficult task in the area of clinical data analysis, is predicting a cardiac disease. Hence in order to solve these issues, Artificial Neural Network with Spider Monkey Optimization Algorithm for Cardio Vascular Disease Prediction is presented in this work. In a recent addition to the group of swarm intelligence-based optimization methods is the Spider Monkey optimization (SMO) algorithm. The combination of SMO and BSA (Bird Swarm optimization Algorithm) namely SMBS (Spider Monkey- Bird Swarm) will be used along with ANN (Artificial Neural Network) to predict the CVD (Cardio Vascular Disease. This System predicts the possibilities (i.e. presence, absence) of CVD. Parameters of prediction accuracy, precision, Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) are used to assess the performance of the presented approach. This approach will achieve better results than earlier prediction models.
Keywords:	Cardio-Vascular Disease (CVD), Prediction, Artificial Neural Network (ANN), Spider Monkey Optimization (SMO), Machine Learning
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	CUSTOMIZING COMMONALITIES GROUNDED INTERNET SERVICE RECOMMENDER SYSTEM USING COLLABORATIVE FILTERING
Author:	K. SELVARANI , Dr. A. ANNADHASON
Abstract:	In the realm of the Internet, Recommender Systems (RS) play a pivotal role in enhancing data retrieval techniques, thereby optimizing the utilization of online data. These systems offer tailored recommendations for items or services to end-users, facilitating well-informed decisions. This paper explores methodologies, specifically the Pearson Correlation-Coefficient (PCC) method in conjunction with Collaborative Filtering, and the Novel Recovery-Collaborative Filtering (NRCF) method, to identify suitable web services. These methods incorporate algorithms for similarity measurement and computation within the domain of Web service recommendation systems. In contrast to existing approaches, this study introduces a novel composite clustering technique that bolsters the accuracy of similarity measurement through precise prediction. The primary goal is to enhance the efficiency of web service recommendations. The evaluation involves subjecting the PCC method to this innovative composite clustering technique, with results presented through comparative analysis.
Keywords:	Customization, Commonalities, Internet Service, Recommender System, Collaborative Filtering
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	ENHANCING LEGAL CASE RETRIEVAL IN MOROCCAN INSOLVENCY LAW USING ONTOLOGY
Author:	KAOUTAR BELHOUCINE, NADIA ZAME, MOHAMMED MOURCHID , ABDELAZIZ MOULOUDI
Abstract:	Retrieving relevant legal cases from digital publications of Moroccan jurisprudence remains a challenge due to the complex nature of the information and the vastness of the database. This research introduces an enhanced Case-Based Retrieval (CBR) system, leveraging ontologies to streamline the retrieval process. The hierarchical structure and logical reasoning capabilities of ontologies enable a comprehensive understanding of semantic relationships. Implementing our system, which combines ontologies with a CBR approach, has demonstrated over 80% accuracy across three legal datasets. This innovation significantly improves the precision of legal case retrieval for Arabic Moroccan jurisprudence.
Keywords:	Legal Retrieval; Moroccan Insolvency Law; Ontologies Building; Case Based Retrieval; Arabic Text.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	TRUST MONITORING FRAMEWORK FOR EFFECTIVE COOPERATION IN SATELLITE ASSISTED UAVS ENABLED VANETS
Author:	F. AL-DOLAIMY, OSAMA YASEEN M. AL-RAWI, RABEI RAAD ALI, AHMED H. R. ALKHAYYAT, FATIMA HASHIM ABBAS, ALI ALSALAMY
Abstract:	Vehicular Ad hoc Network (VANETs) is high-speed networks, and it is combined with Unmanned Aerial Vehicles (UAVs) to build effective communication among vehicles. The communication modules of UAVs assisted VANETs vehicles to roadside units (RSUs), vehicles to UAVs and UAVs to UAVs. And it gets further enabled with satellite and the additional module like UAVs to satellite is included with it. Through UAVs the vehicles are monitored, and it becomes eligible to transfer highly confidential information. So, it is essential to improve the trustworthiness of the vehicles to communicate. For this purpose, Trust Monitoring Framework for Satellite Assisted UAVs Enabled VANETs (TMF-SAUAVs) is proposed in this paper. This method includes two segments such as trust evaluation and trust management. Through trust evaluation, the trust values of vehicles are calculated using direct trust, indirect trust and comprehensive trust calculation. Through trust management, the trust history of the vehicles is properly monitored to easily identify honest vehicles. This method greatly increases the efficiency of the Satellite Assisted UAVs Enabled VANETs. Using NS2 and SUMO, the simulation is run. Energy efficiency, packet delivery rate, end-to-end delay, and routing overhead are the factors considered while evaluating value. The outcomes are contrasted with earlier methods like JRT-UAVs and PDO-UAVs. The results show that, as compared to earlier efforts, the suggested TMF-SAUAVs approach achieved high efficiency and delivery rate as well as lower latency and routing overhead.
Keywords:	Vehicular Ad hoc Network, Unmanned Aerial Vehicles, Satellite Assisted, TMF-SAUAVs.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text

Title:	HBESDM-DLD: A SECURE BLOCKCHAIN-BASED MEDICAL DATA MANAGEMENT WITH DEEP LEARNING-BASED DIAGNOSTIC MODEL
Author:	Mr. SUNIL KUMAR RM, Dr A. JAYACHANDRAN
Abstract:	The healthcare sector has witnessed significant growth in electronic health records (EHRs) generation. While the EHR system provides data owners with control over their data and sharing permissions, the vast amount of healthcare data poses challenges in ensuring security and accurate diagnosis. This paper introduces the HBESDM-DLD model, a novel approach that employs blockchain technology and deep learning for secure medical data management and diagnosis the suggested paradigm contains steps for diagnostics, hyper ledger-based safe data management, optimal key generation, and encryption. Users can manage data access using the model permits hospital authorities to read and write data, and notifies emergency contacts. To enhance security, the elliptical curve cryptography technique is utilized for encryption, with the arithmetic optimization algorithm (AOA) applied for efficient key generation. Multi-channel hyper ledger blockchain is used for sharing medical data, storing patient visit data, and recording EHR links in external databases. Finally, the Spiking Neural Network (SNN)- Following data decryption, an evidence-based diagnostic approach is used to identify disorders. Medical benchmark datasets are used to assess the performance of the HBESDM-DLD model, and better performance than current techniques is shown.
Keywords:	Encrypted Health Record, Encryption, Unauthorized Access, Decryption, Medical Data Management, Optimal Solution
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2023 -- Vol. 101. No. 17-- 2023
Full Text