Saleh Naif Almuayqil

The phenomenon of cyberbullying has emerged as a critical challenge in the digital landscape which poses detrimental effects on individuals and broader societal frameworks. A viable approach to addressing this pervasive issue lies in the accurate identification of cyberbullying within social media as it constitutes a substantial portion of digital interactions. State-of-the-art solutions predominantly rely on pre-trained language models and machine learning algorithms; however, these methods are often associated with substantial computational overheads and the development of advanced cyberbullying detection algorithms remains limited. This paper presents a unique framework that integrates sentiment analysis with machine learning algorithms to enhance the detection of cyberbullying on social media. Sentiment analysis explores the linguistic and emotional characteristics of cyberbullying messages which helps the framework to identify key sentiment indicators that differentiate harmful interactions from benign ones. Furthermore, we provide the most optimal text preprocessing steps which are ordered in a way that improves text quality for cyberbullying detection. These steps ensure high-quality input data for machine learning models which significantly enhances their performance. Additionally, we tackle the challenge of penta-class imbalanced data by incorporating resampling within the framework without causing bias. We employ several machine learning algorithms within the framework, the Extra Tree classifier achieved the best results with an accuracy of 95.38% and F1 score of 0.95. The results demonstrate that the integration of sentiment analysis significantly improves classification accuracy compared to conventional methods for the task of cyberbullying detection.

This study presents KidneyNet, an innovative computer-aided diagnosis (CAD) system designed to identify chronic kidney diseases (CKDs), such as kidney stones, cysts, and tumors, in CT scans. KidneyNet utilizes a convolutional neural network (CNN) structure consisting of eight convolutional layers, three pooling layers, a flattening layer, and two fully connected layers. Small filters enhance computational efficiency by reducing the number of parameters and minimizing the risk of overfitting compared to larger filters. The model captures more complex and abstract features as data move through the layers. The initial layers identify basic patterns, while the deeper layers focus on more intricate representations. KidneyNet aims to enhance the efficiency and accuracy of kidney disease diagnosis. Additionally, the model incorporates the gradient-weighted class activation mapping (Grad-CAM) algorithm, which helps to pinpoint affected areas in the scans. This feature improves interpretability, allowing clinicians to identify which regions the model deemed significant for detecting abnormalities such as tumors, cysts, or stones. Through extensive testing on a CT kidney dataset, KidneyNet demonstrated impressive performance metrics, with 99.88% accuracy, 99.92% specificity, 99.76% sensitivity, 99.58% precision, and an F1 score of 99.67%, outperforming existing models. This approach alleviates the diagnostic burden on radiologists and promotes early detection, potentially saving lives. This study highlights the critical role of advanced imaging analysis in addressing kidney conditions and emphasizes KidneyNet’s capability to deliver precise and cost-effective diagnoses.

Sentiment analysis (SA) is a popular method for obtaining relevant and subjective information from textual information. Multimedia material, such as photos, text, videos, and audio, has lately gained popularity as a way for users to contribute their views on social networking platforms. While sentiment analysis of such material is helpful for various reasons, it is sometimes seen as challenging since these messages are frequently brief, unstructured, noisy, and contain linguistic inconsistencies. Twitter is among the most prominent social media tools for expressing public opinion about various events. However, determining people’s feelings can be tricky since researchers must consider multiple factors. Much of the previous research on sentiment analysis deals with dual or triple-class analysis while using older language modeling techniques. Furthermore, penta-class classification tasks have not been addressed as much. To deal with the challenge, we present a transformer-based model called BertSent that uses ordered preprocessing steps combined with transformer-based tokenization and optimization to get the best sentiment analysis results focused on dealing with limited data. Moreover, our framework handles the challenge of penta-class classification of tweets, and to that end, we combine many preprocessing techniques to fine-tune our BERT-based model. We employ resampling techniques to balance the data which improves model generalization and performance to address class imbalance issues in the penta-class setup. For that purpose, we incorporate both oversampling and undersampling to tackle the challenge of class imbalance when dealing with the penta-class classification problem. Moreover, this article also compares the performance of the transformer-based model against a variety of deep learning-based models, including bi-directional models. The experimentations and results support our model’s remarkable performance considering the limited data and penta-class classification challenge. The results provide an interesting perspective as both undersampling and oversampling provide similar results. BertSent model combined with oversampling provides slightly better performance with 75.3% test accuracy in comparison to undersampling which resulted in 75.1% accuracy.

The security of stego-images is a crucial foundation for analyzing steganography algorithms. Recently, steganography has made significant strides in ongoing conflicts with steganalysis. In order to increase the security of stego-images, steganography must be able to evade detection using steganalysis methods. Secret information is typically hidden using traditional embedding-based steganography, which inevitably leaves traces of the modifications that can be found using more sophisticated machine-learning-based steganalysis techniques. Steganography without embedding (SWE) outperforms machine-learning-based steganalysis techniques because it does not require alteration of the data of the cover image. A novel image SWE method based on deep convolutional generative adversarial networks (GANs) is proposed to synthesize stego-images led by embedded text. The variational autoencoder (VAE) in the GAN model is utilized to synthesize the stego-image, based on interpolating the secret text in a continuous variable representation of the cover image. To further improve the framework’s performance and shorten processing times, the whale optimization algorithm (WOA) is used to identify the optimal VAE structure. When creating a stego-image, no embedding or modification procedures are required, and after training, a different convolutional neural network (CNN) known as the extractor can correctly extract the data from the image. The experimental results revealed that this approach has the advantages of evading detection using innovative deep learning (DL) steganalysis architecture and accurate information extraction.

Cancer diagnosis and treatment depend on accurate cancer-type prediction. A prediction model can infer significant cancer features (genes). Gene expression is among the most frequently used features in cancer detection. Deep Learning (DL) architectures, which demonstrate cutting-edge performance in many disciplines, are not appropriate for the gene expression data since it contains a few samples with thousands of features. This study presents an approach that applies three feature selection techniques (Lasso, Random Forest, and Chi-Square) on gene expression data obtained from Pan-Cancer Atlas through the TCGA Firehose Data using R statistical software version 4.2.2. We calculated the feature importance of each selection method. Then we calculated the mean of the feature importance to determine the threshold for selecting the most relevant features. We constructed five models with a simple convolutional neural networks (CNNs) architecture, which are trained using the selected features and then selected the winning model. The winning model achieved a precision of 94.11%, a recall of 94.26%, an F1-score of 94.14%, and an accuracy of 96.16% on a test set.

In the face of the COVID-19 pandemic, many studies have been undertaken to provide assistive recommendations to patients to help overcome the burden of the expected shortage in clinicians. Thus, this study focused on diagnosing the COVID-19 virus using a set of fine-tuned deep learning models to overcome the latency in virus checkups. Five recent deep learning algorithms (EfficientB0, VGG-19, DenseNet121, EfficientB7, and MobileNetV2) were utilized to label both CT scan and chest X-ray images as positive or negative for COVID-19. The experimental results showed the superiority of the proposed method compared to state-of-the-art methods in terms of precision, sensitivity, specificity, F1 score, accuracy, and data access time.

Recently, smart home devices have been widely used to assist and facilitate the lives of human beings. Human activity recognition (HAR) aims to identify human activities using sensors in smartphones. Therefore, it can be employed in many applications, such as remote health monitoring for disabled and elderly people. This paper proposes a granular computing-based approach to classifying human activities using wearable sensing devices. The approach has two main phases: feature selection and classification. In the feature selection phase, the approach attempts to remove redundant and irrelevant attributes. At the same time, the classification phase makes use of granular computing concepts to build the granules and find the relationships between granules at different levels. To evaluate the approach, we applied the dataset to five famous machine learning models. For the comparative evaluation, we also tested other well-known machine learning methods. The experimental results presented in this paper show that the approach outperformed common traditional classifiers in terms of classification precision recall, f-measure, and MCC for most recognized human activities by approximately 97.3%, 94%, 95.5%, and 94.8%, respectively. However, in terms of processing time, it performs comparably.

Cancer is a complicated global health concern with a significant fatality rate. Breast cancer is among the leading causes of mortality each year. Advancements in prognoses have been progressively based primarily on the expression of genes, offering insight into robust and appropriate healthcare decisions, owing to the fast growth of advanced throughput sequencing techniques and the use of various deep learning approaches that have arisen in the past few years. Diagnostic-imaging disease indicators such as breast density and tissue texture are widely used by physicians and automated technology. The effective and specific identification of cancer risk presence can be used to inform tailored screening and preventive decisions. For several classifications and prediction applications, such as breast imaging, deep learning has increasingly emerged as an effective method. We present a deep learning model approach for predicting breast cancer risk primarily on this foundation. The proposed methodology is based on transfer learning using the InceptionResNetV2 deep learning model. Our experimental work on a breast cancer dataset demonstrates high model performance, with 91% accuracy. The proposed model includes risk markers that are used to improve breast cancer risk assessment scores and presents promising results compared to existing approaches. Deep learning models include risk markers that are used to improve accuracy scores. This article depicts breast cancer risk indicators, defines the proper usage, features, and limits of each risk forecasting model, and examines the increasing role of deep learning (DL) in risk detection. The proposed model could potentially be used to automate various types of medical imaging techniques.

Adoption of Ambient Assisted Living (AAL) technologies for geriatric healthcare is suboptimal. This study aims to present the AAL Adoption Diamond Framework, encompassing a set of key enablers/barriers as factors, and describe our approach to developing this framework. A systematic literature review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. SCOPUS, IEEE Xplore, PubMed, ProQuest, Science Direct, ACM Digital Library, SpringerLink, Wiley Online Library and grey literature were searched. Thematic analysis was performed to identify factors reported or perceived to be important for adopting AAL technologies. Of 3717 studies initially retrieved, 109 were thoroughly screened and 52 met our inclusion criteria. Nineteen unique technology adoption factors were identified. The most common factor was privacy (50%) whereas data accuracy and affordability were the least common factors (4%). The highest number of factors found per a given study was eleven whereas the average number of factors across all studies included in our sample was four (mean = 3.9). We formed an AAL technology adoption framework based on the retrieved information and named it the AAL Adoption Diamond Framework. This holistic framework was formed by organising the identified technology adoption factors into four key dimensions: Human, Technology, Business, and Organisation. To conclude, the AAL Adoption Diamond Framework is holistic in term of recognizing key factors for the adoption of AAL technologies, and novel and unmatched in term of structuring them into four overarching themes or dimensions, bringing together the individual and the systemic factors evolving around the adoption of AAL technology. This framework is useful for stakeholders (e.g., decision-makers, healthcare providers, and caregivers) to adopt and implement AAL technologies.

According to medical reports and statistics, skin diseases have millions of victims worldwide. These diseases might affect the health and life of patients and increase the costs of healthcare services. Delays in diagnosing such diseases make it difficult to overcome the consequences of these types of disease. Usually, diagnosis is performed using dermoscopic images, where specialists utilize certain measures to produce the results. This approach to diagnosis faces multiple disadvantages, such as overlapping infectious and inflammatory skin diseases and high levels of visual diversity, obstructing accurate diagnosis. Therefore, this article uses medical image analysis and artificial intelligence to present an automatic diagnosis system of different skin lesion categories using dermoscopic images. The addressed diseases are actinic keratoses (solar keratoses), benign keratosis (BKL), melanocytic nevi (NV), basal cell carcinoma (BCC), dermatofibroma (DF), melanoma (MEL), and vascular skin lesions (VASC). The proposed system consists of four main steps: (i) preprocessing the input raw image data and metadata; (ii) feature extraction using six pre-trained deep learning models (i.e., VGG19, InceptionV3, ResNet50, DenseNet201, and Xception); (iii) features concatenation; and (iv) classification/diagnosis using machine learning techniques. The evaluation results showed an average accuracy, sensitivity, specificity, precision, and disc similarity coefficient (DSC) of around 99.94%, 91.48%, 98.82%, 97.01%, and 94.00%, respectively.