Open Access Journals

Explore All Journals...

Recent Articles >> More

Design of Congestion Prediction Model on Network Towers using Rough Set Theory for High Speed Low Latency Communication

By D. Priyanka Y.K. Sundara Krishna

DOI: https://doi.org/10.5815/ijcnis.2025.02.01, Pub. Date: 8 Apr. 2025

Wireless communication for data and a variety of wireless interacted devices have increased dramatically in the past few years. Millimeter wave (mmWave) technology can serve the primary objectives of 5G networks, which include high data throughput and low latency. But mmWave signals for communications lacking substantial diffraction and are consequently more susceptible to obstruction by environmental physical objects, which could cause communication lines to be disrupted and congestion takes place. Wireless data transmission suffers from blockages and path loss, causes high latency as well as reduces the data transmission speed and degrades in quality performance. To overcome the limitations, Rough Set Theory with hypertuned SVM is implemented and designed the congestion prediction model based on the behaviour of network towers for low latency and high-speed data transmission. The data from the different towers is initially collected and created as a dataset. Super MICE is a technique to replace the missing data. Then, the Rough Set Theory is utilized to cluster the data into equivalent classes based on the behaviour of 5G, 4G and 3G wireless network. Hypertuned SVM with a Gazelle optimization algorithm is applied to predict the congestion level by accurately selecting the hyperparameter. By employing performance metrics, the proposed approach is examined and contrasted with existing techniques. The evaluation of performance measurements for the proposed method includes informedness attained as 91%, Adjusted Rand Index obtained value as 0.83, Jaccard as 0.737. Accuracy, precision, sensitivity, error, F1_score, and NPV are also achieved at 93%, 92%, 94%, 7%, 92%, and 90%, respectively. According to this evaluation, the proposed model is superior to perform than the earlier used existing methods.  

[...] Read more.
Identifying Dark Web Hidden Services with Novel Image Classes Using CNN and Quantum Transfer Learning

By Ashwini Dalvi Soham Bhoir Akansha Singh Irfan Siddavatam Sunil Bhirud

DOI: https://doi.org/10.5815/ijeme.2025.02.05, Pub. Date: 8 Apr. 2025

The dark web is an overwhelming and mysterious place that comprises hidden services. Dark web hidden services contain illegal or offensive content. Hidden services are not accessible through regular search engines or browsers and can only be accessed via specific software. The proposed work aims to identify these hidden services by analyzing their associated images and text data. Doing so, one can better understand the types of activities on the dark web and what kind of content is available. First, a dark web crawler is developed to collect dark web services. Images are then manually classified into four categories: Cards, Devices, Hackers, and Money. Next, preprocessing the collected dataset removed irrelevant images, and a Convolutional Neural Network (CNN) was trained to identify new dark web image classes. Finally, quantum Transfer Learning (QTL) improved the model’s performance. The proposed work goes beyond conventional methods of categorizing datasets by including new categories of image classes of dark web hidden services that have not been considered before. Also, the work examines image data and related text to establish a strong correlation between them. The proposed approach will provide insights into the dark web hidden service by confirming the relationship between the image and text data of the respective hidden-services.

[...] Read more.
Advancing Blood Cancer Diagnostics: A Comprehensive Deep Learning Framework for Automated and Precise Classification

By Md. Samrat Ali Abu Kawser Md. Showrov Hossen

DOI: https://doi.org/10.5815/ijem.2025.02.04, Pub. Date: 8 Apr. 2025

A vital component of patient care is the diagnosis of blood cancer, which necessitates prompt and correct classification for efficient treatment planning. The limitations of subjectivity and different levels of skill in manual classification methods highlight the need for automated systems. This study improves blood cancer cell identification and categorization by utilizing deep learning, a subset of artificial intelligence. Our technique uses bespoke U-Net, MobileNet V2, and VGG-16, powerful neural networks to address problems with manual classification. For the purposes biomedical image segmentation U-Net architecture is used, MobileNet V2 is used for lightweight neural network model design and VGG-16 is used for image classification. A hand-picked dataset from Taleqani Hospital in Iran is used for the rigorous training, validation, and testing of the suggested models. The dataset is refined using denoising, augmentation, and linear normalisation, which improves model adaptability. The results show that the MobileNet V2 model outperforms related studies in terms of accuracy (97.42%) when it comes to identifying and categorizing blast cells from acute lymphoblastic leukemia. This work offers a fresh approach that adds to artificial intelligence's potentially revolutionary potential in medical diagnosis.

[...] Read more.
A Novel Approach for Enhancing COVID-19 Diagnosis Accuracy through Graph Neural Networks Using Respiratory Sound Data

By Nagaraju Sonti Rukmini M. S. S. Venkatesh Munagala

DOI: https://doi.org/10.5815/ijigsp.2025.02.03, Pub. Date: 8 Apr. 2025

This research presents a groundbreaking method using graph neural networks (GNN) for the accurate identification of COVID-19 through the analysis of respiratory sounds. The method utilizes advanced signal processing and machine learning techniques, including Fast Fourier Transforms (FFTs), Mel-spectrograms, and GNN methodology. FFTs are used as a preprocessing step to convert raw respiratory sound signals into frequency-domain representations, enhancing signal quality and isolating informative acoustic patterns. Mel-spectrograms are used to extract essential feature vectors for diagnostic classification, enhancing the model's ability to discern subtle patterns indicative of COVID-19 infection.
The GNN methodology feeds preprocessed audio features into a graph neural network architecture, which excels at capturing complex relationships and dependencies within data by modeling them as graphs. In this context, respiratory sound data is represented as a graph, with nodes corresponding to specific audio features and edges representing relationships between them. The GNN effectively learns to propagate information across the graph, enabling it to identify meaningful patterns indicative of COVID-19 infection. The research findings show that GNN surpasses convolutional neural network (CNN) in terms of accuracy, precision, recall, and F1 score, indicating significant progress in the application of GNN in medical diagnostics. The study provides a comprehensive examination of the possibilities of using advanced neural network techniques to transform disease detection and diagnosis, with a validation accuracy of up to 97% under rigorous constraints.

[...] Read more.
Agile Methodology of Information Engineering for Semantic Annotations Categorization and Creation in Scientific Articles Based on NLP and Machine Learning Methods

By Danylo Levkivskyi Victoria Vysotska Lyubomyr Chyrun Yuriy Ushenko Dmytro Uhryn Cennuo Hu

DOI: https://doi.org/10.5815/ijieeb.2025.02.01, Pub. Date: 8 Apr. 2025

Research devoted to the categorization and creation of semantic annotations for scientific articles stands out as an essential direction of development in the context of the growing volume of scientific literature. The application of machine learning and natural language processing in this field allows you to effectively organize and provide access to scientific information. The article discusses methods of automatic annotation of texts. Based on the review, the use of the constraint propagation model is proposed to improve the technique of text relationship maps. The developed software system is aimed at automating the process of analysis and categorization of scientific materials, which opens the way to improving the speed and accuracy of searching for the necessary information for researchers. The use of advanced machine learning models, such as roBERTa and RAG, ensures the highest quality of data processing and creation of semantic annotations. The accuracy of predicting article categories after improving the model reached 88%. The novelty of the approach is the combination of categorization and semantic annotation to increase the convenience and speed of searching for scientific information. The software system opens up opportunities for future expansion and improvement through the use of advanced technologies and machine learning models. This study is noted for its relevance, originality of approach and potential for practical application in the field of scientific research and development of science as a whole. The proposed approach contributes to the development of the Information Engineering and Electronic Business industry through the following key aspects: automation of categorization and annotation of scientific articles, improving the accuracy of information search, increasing the efficiency of scientific research, and the flexibility and scalability of the solution.

[...] Read more.
A Survey of Techniques for Improving Information Retrieval through Query Expansion

By Surabhi Solanki Seema Verma Sachin Kumar

DOI: https://doi.org/10.5815/ijitcs.2025.02.07, Pub. Date: 8 Apr. 2025

This paper presents a comprehensive survey of QE techniques in IR. Core techniques, employed data sources, and methodologies used in the process of query expansion are discussed. The output study highlights four main steps concerned with expanding queries: steps related to preprocessing of data sources and term extraction, calculation of weights and ranking of terms, selection of terms, and finally expansion. The most important findings are that only effective text normalization and removal of stopwords provide a real platform for performing QE. The introduction of contextually relevant terms significantly enhanced relevance feedback and thesaurus-based WordNet expansion techniques. They have been shown to significantly improve retrieval effectiveness as has been realized from various experiments conducted over years now. It also uses the manual query expansion techniques and discusses several automated ways in order to improve retrieval effectiveness. This work, by reviewing the related literature and methodologies, gives an overview of how the techniques of query expansion have been evolving with time and achieved better results in IR systems. The survey offers a valuable resource for researchers and practitioners in information retrieval, shedding light on the advancements, challenges, and future directions in query expansion research.

[...] Read more.
Unveiling Hidden Patterns: A Deep Learning Framework Utilizing PCA for Fraudulent Scheme Detection in Supply Chain Analytics

By Kowshik Sankar Roy Pritom Biswas Udas Bashirul Alam Koushik Paul

DOI: https://doi.org/10.5815/ijisa.2025.02.02, Pub. Date: 8 Apr. 2025

Supply chain fraud, a persistent issue over the decades, has seen a significant rise in both prevalence and sophistication in recent years. In the current landscape of supply chain management, the increasing complexity of fraudulent activities demands the use of advanced analytical tools. Despite numerous studies in this domain, many have fallen short in exploring the full extent of recent developments. Thus, this paper introduces an innovative deep learning-based classification model specifically designed for fraud detection in supply chain analytics. To enhance the model's performance, hyperparameters are fine-tuned using Bayesian optimization techniques. To manage the challenges posed by high-dimensional data, Principal Component Analysis (PCA) is applied to streamline data dimensions. In order to address class imbalance, the SMOTE technique has been employed for oversampling the minority class of the dataset. The model's robustness is validated through evaluation on the well-established 'DataCo smart supply chain for big data analysis' dataset, yielding impressive results. The proposed approach achieves a 94.71% fraud detection rate and an overall accuracy of 99.42%. Comparative analysis with various other models highlights the significant improvements in fraud transaction detection achieved by this approach. While the model demonstrates high accuracy, it may not be directly transferable to more diverse or real-world datasets. As part of future work, the model can be tested on more varied datasets and refined to enhance generalizability, better aligning it with real-world scenarios. This will include addressing potential overfitting to the specific dataset used and ensuring further validation across different environments to confirm the model's robustness and generalizability.

[...] Read more.
On the Noisy Four-Parameter Fisher‟s Z-Distribution of Bayesian Mixture Autoregressive (FZBMAR) Process via Mode as a Stable Location Parameter

By Rasaki Olawale Olanrewaju Sodiq Adejare Olanrewaju

DOI: https://doi.org/10.5815/ijmsc.2025.01.05, Pub. Date: 8 Apr. 2025

This paper aims at providing in-depth refinement to switching time-variant autoregressive processes via the mode as a stable location parameter in adopted noisy Fisher’s z-distribution that was impelled in a Bayesian setting. Explicitly, a four-parameter Fisher’s z-distribution of Bayesian Mixture Autoregressive (FZBMAR) process was proposed to congruous  k-mixture components of Fisher’s z-switching mixture autoregressive processes that was based on shifting number of modes in the marginal density of any switching time-variant series of interest. The proposed FZBMAR process was not only used to seize what is term “most likely mode value” of the present conditional modal distribution given the immediate past but was also used to capture the conditional modal distribution of the observations given the immediate past that can either be perceived as an asymmetric or symmetric distributed varieties. The proposed FZBMAR process was compared with the existing Student-t Mixture Autoregressive (StMAR) and Gaussian Mixture Autoregressive (GMAR) processes with the demonstration of monthly average share prices (stock prices) of sixteen (16) swaying European economies. Based on the findings, the FZBMAR process outperformed the existing StMAR and GMAR processes in explaining the sixteen (16) swaying European economies share prices via a minimum Pareto-Smoothed Important Sampling Leave-One-Out Cross-Validation (PSIS-LOO) error process performance in comparison with AIC, HQIC by the latters. The same singly truncated student-t prior distribution was adopted for the noisy adoption of Fisher’s z hyper-parameters and the embedded autoregressive coefficients in the proposed FZBMAR process; such that their resulting posterior distributions gave the same singly truncated student-t distribution (conjugate) with an embedded Gamma variate.

[...] Read more.
Assessing Student Academic Performance with Fuzzy Expert System

By Bhupendra Kumar Pathak

DOI: https://doi.org/10.5815/ijmecs.2025.02.05, Pub. Date: 8 Apr. 2025

Nowadays, higher education institutions and universities are facing a competitive environment for enhancing the quality of students to achieve extensive knowledge with critical thinking skills and a good personality for better employment in the industry. Universities and other higher education establishments ensure that students overcome the obstacles in these cutthroat environments. In order to do this, it is necessary to analyze the academic performance of each student by determining their strengths and weaknesses. A fuzzy expert system (FES) is used in this study to evaluate student’s academic performance. This FES uses fuzzy logic to classify each student’s performance based on a variety of linguistic factors. It classifies each student’s performance by considering various linguistic factors using fuzzy logic. For this purpose, seven significant input factors have been taken into account which is Stress, Motivation, Confidence, Parent’s support & Availability, Self study hours, Punctuality, and Friend circle. Several defuzzification techniques are applied in order to examine student performance using the FES & generate more precise and measurable results. These findings could aid colleges and other educational establishments in determining the right variables that influence student’s academic performance. Additionally, a comparison of various Mamdani fuzzy defuzzification techniques, including the centroid, bisector, and mean of maxima methods, is provided in this study. After comparing all three techniques by taking different scenarios of all the external factors, it has been concluded that all of them are performing equally.

[...] Read more.
Novel Machine Learning Approaches for Identifying Attacks in IoT-based Smart Home Environment

By Oyelakin A. M. Sanni S. A. Adegbola I. A. Salau-Ibrahim T. T. Bakare-Busari Z. M. Saka B. A.

DOI: https://doi.org/10.5815/ijwmt.2025.02.04, Pub. Date: 8 Apr. 2025

Attackers keep launching different attacks on computer networks. Signature-based and Machine Learning (ML)-based techniques have been used to build models for promptly identifying these attacks in networks. However, ML-based approaches are more popular than their counterparts because of their ability to detect zero-day attacks.  In the Internet of Things (IoT), devices are interconnected and this called for the need to guide such networks against intrusions. This study aims at building effective ML models from a recently released IoT-based Smart Home dataset. The study revealed patterns and characteristics of the IoT dataset, pre-processed it and then selected discriminant features using Binary Bat Algorithm (BBA). The pre-processing of the Smart Home IoT dataset for the study was carried out based on the issues identified during the exploratory analyses. The experimental evaluation carried out revealed that all the learning algorithms achieved promising classification results. For instance, Decision Trees recorded 98.60% accuracy, KNN produced 99.60% accuracy while Random Forest (RF) and AdaBoost-based models recorded 100.00% and 99.91% respectively. In all other metrics, RF-based attack classification model slightly recorded the best results. The study concluded that the EDA, innovative data pre-processing, BBA-based feature selection improved the classification performances of the ML approaches used in this study.

[...] Read more.

More...