Recent Advances in Computer Science and Communications - Online First

Cybersecurity requirements for ensuring data security during research manuscript transit on the journal website require continuous improvement and adherence to best practices. Research data loss can have significant negative consequences across multiple dimensions including time and financial loss. The present research investigates security vulnerabilities during the real-time transit of manuscripts on a journal website.

Procedure: Website Access: The journal website was accessed, and manuscript components (main manuscript, figures, tables, graphical abstract, funding sources, suggested reviewer, and cover letter) were uploaded.

Operating system: Kali Linux, designed for penetration testing and security auditing was used.

Tools and software: Nmap (Version 7.95-2) for network discovery and security auditing. Nikto (2.5.0) for web server vulnerability scanning, and Tor (13.0.13) to anonymize web activities. Firefox (127.0.2) as the web browser, and VMware Workstation with Kali Rolling (2023.2 in a virtual environment.

Testing phase: Initial upload of the manuscript and supplementary materials. Upload of figures, tables, and graphical abstract. Inclusion of funding sources, suggested reviewers, and cover letter.

Data Collection and Analysis: Network traffic and potential vulnerabilities were monitored on Nmap, Nikto, and Tor.

Activities were conducted in the virtual environment of VMware Workstation for controlled and replicable setup.

Output measures: Identified and documented potential security gap or vulnerabilities leading to data theft during manuscript transit.

An Nmap scan of XXXXXXXX.com (IP: yyyyyyyyyyy) revealed six open ports: 80 (HTTP Apache), 443 (SSL/SMTP Exim), 587 (SMTP Exim), 993 (IMAPS), and 995 (POP3S). each server showed potential vulnerabilities. The scan took 86.15 seconds.

The results demonstrate a high risk of exposing sensitive information due to open ports, the presence of potentially outdated services, and the possibility of incomplete detection due to filtered ports pose a high risk of sensitive data during manuscript transit on the website of the journal.

Considering the singularity of collaborative filtering algorithms in recommending learning resources and the problem that existing knowledge graph convolutional networks cannot deeply mine the neighbourhood information of learning resource nodes in application scenarios with less neighbourhood information, a multi-domain fusion convolutional network learning resource recommendation model based on knowledge graphs is proposed here.

This study aimed to improve the accuracy and personalization of recommendations of filtering algorithms.

First, the model mapped learner nodes, learning resources, and their neighbour nodes into low-dimensional dense vectors. Second, the multi-domain fusion layer and the multi-domain aggregator were used to obtain the fused multi-domain learning resource vector. Finally, the learner vector and the multi-domain learning resource vector were fed into the prediction layer to calculate the interaction probability.

To verify the effectiveness of the algorithm, we conducted a comparative experiment using the publicly available datasets MOOPer and MOOCCubeX. The experimental results showed that the proposed model outperformed baseline models, such as CKE, MKR, KGCN, DEKGCN, and KGIN, in terms of evaluation metrics, such as AUC, ACC, and F1. At the same time, when the neighborhood information was limited, the AUC, ACC, and F1 values of the proposed model still maintained the optimal value.

Compared to the optimal baseline model, the effectiveness of the proposed model has been proven.

With the rapid development of society, motor vehicles have become one of the main means of transportation. However, as the number of motor vehicles continues to increase, traffic safety accidents also continue to appear, bringing serious threats to people's lives, and property safety. Fatigue driving is one of the important causes of traffic safety accidents.

To address this problem, a target detection algorithm called VA-YOLO is designed to improve the speed and accuracy of facial recognition for fatigue checking. The algorithm employs a lightweight backbone network, VanillaNet, instead of the traditional backbone network, which reduces the computational and parametric quantities of the model. The SE attention mechanism is also introduced to enhance the model's attention to the target features, which further improves the accuracy of target detection. Finally, in terms of the bounding box regression loss function, the SIoU loss function is used to reduce the error.

The experimental results show that, compared toYolov8n, the VA-YOLO algorithm improves the accuracy by 1.3% while the number of parameters decreases by 30%.

This shows that the VA-YOLO algorithm has a significant advantage in realizing the balance between the number of parameters and accuracy, which is important for improving the speed and accuracy of fatigue driving detection.

Unbalanced datasets present a significant challenge in machine learning, often leading to biased models that favor the majority class. Recent oversampling techniques like SMOTE, Borderline SMOTE, and ADASYN attempt to mitigate these issues. This study investigates these techniques in conjunction with machine learning models like SVM, Decision Tree, and Logistic Regression. The results reveal critical challenges such as noise amplification and overfitting, which we address by refining the oversampling approaches to improve model performance and generalization.

In order to address this challenge of unbalanced datasets, the minority class is oversampled to accommodate the majority class. Oversampling techniques such SMOTE (Synthetic Minority Oversampling Technique), Borderline SMOTE and ADASYN (Adaptive Synthetic Sampling) are used in this work.

To perform the comprehensive analysis of various oversampling methods for taking acre of class imbalance issue using ML methods.

The proposed methodology uses BERT technique which removes the pre-processing step. Various proposed oversampling techniques in the literature are used for balancing the data, followed by feature extraction followed by text classification using ML algorithms. Experiments are performed using ML classification algorithms like Decision tree (DT), Logistic regression (LR), Support vector machine (SVM) and Random forest (RF) for categorizing the data.

The results show improvement corresponding SVM using Borderline SMOTE, resulting in an accuracy of 71.9% and MCC value of 0.53.

The suggested method assists in the evolution of fairer and more effective ML models by addressing this basic issue of class imbalance.

Deformable image registration is an essential task in medical image analysis. The UNet model, or the model with the U-shaped structure, has been popularly proposed in deep learning-based registration methods. However, they easily lose the important similarity information in the up-sampling stage, and these methods usually ignore the inherent inverse consistency of the transformation between a pair of images. Furthermore, the traditional smoothing constraints used in the existing methods can only partially ensure the folding of the deformation field.

An inverse consistent deformable medical image registration network(ICSANet) based on the inverse consistency constraint and the similarity-based local attention is developed. A new UNet network is constructed by introducing similarity-based local attention to focus on the spatial correspondence in the high-similarity space. A novel inverse consistency constraint is proposed, and the objective function of the new form is presented with the combination of the traditional constraint conditions. Experiment: The performance of the proposed method is compared with the typical registration models, such as the VoxelMorph, PVT, nnFormer, and TransMorph-diff model, on the brain IXI and OASIS datasets.

Experimental results on the brain MRI datasets show that the images can be deformed symmetrically until two distorted images are well matched. The quantitative comparison and visual analysis indicate that the proposed method performs better, and the Dice index can be improved by at least 12% with only 10% parameters.

This paper presents a new medical image registration network, ICSANet. By introducing a similarity attention gate, it accurately captures high-similarity spatial correspondences between source and target images, resulting in better registration performance.

Attention-Deficit Hyperactivity Disorder (ADHD) is a prevalent chronic mental health condition that significantly impacts the psychological and physical well-being of millions of adolescents. Early detection and accurate diagnosis are crucial for effective treatment and mitigating the disorder's adverse effects. Despite extensive research efforts, current methods often fall short in simultaneously accounting for daily motor activity and heart rate variability in ADHD detection.

Addressing these gaps, this paper introduces a histogram-based gradient-boosting classifier for analyzing real-time activity and heart-rate variability data to automate ADHD diagnosis. By extracting twelve key features from the data and selecting the most significant ones with the extra tree model, we evaluate these features using various classifiers, including histogram-based gradient boosting, light gradient boosting machine, extreme gradient boosting, gradient boosting, and adaptive boosting.

The histogram-based gradient-boosting model, validated through ten-fold cross-validation, outperforms other classifiers with an accuracy of 99.12%, an F1 measure of 99.12%, and a sensitivity of 99.13%. Additionally, it achieves a specificity of 99.1%, an AUC of 0.9995, and a minimal FDR of 0.88%. These results demonstrate that the proposed approach offers a highly effective and precise solution for automated ADHD diagnosis.

The implications of these findings suggest that integrating real-time activity and heart-rate variability data into diagnostic processes can significantly enhance the accuracy and efficiency of ADHD assessment, potentially leading to earlier and more reliable diagnoses, improved patient outcomes, and more tailored treatment strategies.

Nowadays, Artificial intelligence and machine learning have emerged as a powerful tool for the analysis of medical images such as MRI scans. This technology holds significant potential to improve diagnostic services and accelerate medical advances by facilitating clinical decision-making.

In this work, we developed a Convolutional Neural Network (CNN) model specifically designed for the classification of medical images. Using a selected database, the model achieved a classification accuracy of 92%. To further improve the performance, we leveraged the pre-trained VGG16 model, which increased the classification accuracy to 100%. Additionally, we preprocessed the MRI images using the Roboflow platform and then developed YOLOv5 models for the detection of tumors, infections, and cancerous lesions.

The results demonstrate a localization accuracy of 50.41% for these medical conditions.

This research highlights the value of AI-driven approaches in enhancing medical image analysis and their potential to support more accurate diagnoses and accelerate advancements in healthcare.