Hazrat Bilal , Yar Muhammad , Inam Ullah , Sahil Garg , Bong Jun Choi , Mohammad Mehedi Hassan
{"title":"Identification and diagnosis of chronic heart disease: A deep learning-based hybrid approach","authors":"Hazrat Bilal , Yar Muhammad , Inam Ullah , Sahil Garg , Bong Jun Choi , Mohammad Mehedi Hassan","doi":"10.1016/j.aej.2025.03.025","DOIUrl":"10.1016/j.aej.2025.03.025","url":null,"abstract":"<div><div>Chronic heart disease has emerged as a challenging issue in the healthcare sector that needs serious attention to save the lives of millions of cardiac patients. The precise diagnosis of this disease in the early stages can reduce the devastating effect it has on human life. To address this issue, this study proposes a hybrid deep learning (DL)-based approach that combines two versatile DL models, namely, bidirectional long-short-term memory (BLSTM) and bidirectional gated recurrent unit (BGRU), resulting in an efficient hybrid DL model named BLSTM-BGRU. The BLSTM part captures long-term relationships between dataset attributes, guaranteeing the preservation of the patient’s historical data, which is essential for forecasting the patient’s health conditions. The BGRU part improves the computing efficiency of the model by lowering the number of trainable parameters and reducing the effect of vanishing gradient problems. The integration of BLSTM and BGRU helps the model to learn the short-term variations and long-range dependencies in heart disease attributes such as heart rate, respiratory rate, etc. The proposed model captures contextual dependency in forward and backward directions, resulting in improved heart disease diagnostic accuracy by learning long-range relationships between attributes and complex sequences. To determine the efficiency of the BLSTM-BGRU model, the MIT-BIH dataset, which consists of five different types of ECG signals, was used. The dataset consists of more normal class samples than the rest of the four classes. Therefore, we used the SMOTE dataset balancing technique to balance the dataset, thereby avoiding the model overfitting problem and improving its efficiency. Alongside the proposed model, we also investigated the performance of four other of the most versatile DL models on both unbalanced and balanced datasets. The proposed model achieved training and testing accuracy of 99.90% and 99.58% on an unbalanced dataset and 99.95% and 99.70%, respectively, on a balanced dataset. The results highlight the importance of the proposed BLSTM-BGRU model using both balanced and unbalanced datasets, showing its significance and versatility for the identification of heart disease, resulting in enhanced heart disease prevention and management.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 470-483"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817527","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Security application of intrusion detection model based on deep learning in english online education","authors":"Xue Li , Yugui Zhang","doi":"10.1016/j.aej.2025.03.051","DOIUrl":"10.1016/j.aej.2025.03.051","url":null,"abstract":"<div><div>Nowadays, the issue of English online education security has become increasingly prominent. The increasing complexity and concealment of cyber-attack cause significant financial losses in English online education application, exacerbating the distrust of teachers and students towards cyberspace. Therefore, this paper proposes a multi scale convolutional neural network based on multi head attention mechanism and hierarchical long short term memory network (MCNN-MHA-HLSTM). This model uses one dimensional convolution to construct a multi scale convolution structure to extract network data feature information of different scales. And it combines multi head attention mechanism to enhance the weight of features related to English online education data, for improving the intrusion detection capability in English online education. Simultaneously it designs a hierarchical long short term memory network (HLSTM) to extract temporal features across multiple temporal hierarchical structure on network data sequences. Finally, the experimental results display that MCNN-MHA-HLSTM can significantly improve the intrusion detection capability of English online education platforms, laying a technical foundation for the operation and sustainable development of English online education security application.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 582-590"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"T_SRNET: A multimodal model based on convolutional neural network for emotional speech enhancement","authors":"Shaoqiang Wang , Lei Feng , Li Zhang","doi":"10.1016/j.aej.2025.03.071","DOIUrl":"10.1016/j.aej.2025.03.071","url":null,"abstract":"<div><div>Speech classification is a technology that can determine the emotional state conveyed by speech. It can support emotion-related applications and improve the human–computer interaction experience. However, the lack of high-quality speech annotation datasets makes it difficult for many models to provide sufficient data for training, resulting in poor model generalization performance. It is necessary to obtain more high-quality speech annotation datasets through the high-precision model. For example, there are many human emotional data in the image dataset that can be utilized to assist in speech emotional information recognition. In this study, a multimodal algorithm T_SRNET is proposed, which can assist speech emotion recognition by extracting image emotion features and converting them into spectrograms. Firstly, the face image data with emotions such as joy and sadness are transformed into the corresponding phonograms by the diffusion model. Secondly, the features can be extracted by using the speech feature extraction network SRNET based on the improved transform structure. Finally, the speech signal features are extracted, and the two features are fused before the decision is made to output the results. After ablation and contrast experiments, the accuracy of CREMA-D and IEMOCAP was improved by 2% and 1% respectively. Also it can be evaluated that the proposed model in this study can correlate image data with speech data, improve the quality of speech data tagging and enhance the performance of speech recognition.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 573-581"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Umesh Gupta , Shubham Kandpal , Hayam Alamro , Mashael M. Asiri , Meshari H. Alanazi (;) , Ali M. Al-Sharafi , Shaymaa Sorour
{"title":"Efficient malware detection using NLP and deep learning model","authors":"Umesh Gupta , Shubham Kandpal , Hayam Alamro , Mashael M. Asiri , Meshari H. Alanazi (;) , Ali M. Al-Sharafi , Shaymaa Sorour","doi":"10.1016/j.aej.2025.03.118","DOIUrl":"10.1016/j.aej.2025.03.118","url":null,"abstract":"<div><div>Malware has emerged as a significant challenge in contemporary society, growing in tandem with technological advancements. Consequently, the classification of malware has become a pressing concern for various services. Conventional malware detection techniques, such as signature matching, are constrained by the dynamic evolution of malware, which limits their adaptability and efficacy. To tackle these issues, this study employs natural language processing (NLP) and deep learning approaches to categorize malware entities as either malicious or benign. The model incorporates image processing by transforming code segments into image pixels, applying convolutional operations, and utilizing advanced deep learning methodologies. Following processing, the model generates a normalized value through the sigmoid function, which is then rounded to yield a binary classification. The results were validated using multiple metrics, including precision and accuracy, to evaluate the model's effectiveness and ensure optimal performance throughout the classification process. The proposed model's performance was assessed on datasets of kernel API calls by the malware. The research highlights that using NLP from the function calls and deep learning techniques for malware classification enhances the accuracy and adaptability of detecting malicious software which overcomes the limitations of traditional signature-based methods. The model delivers encouraging results, presenting a viable solution for effective malware classification. This paper aims to experiment with different variables of a malicious code that are often overlooked while analysing a malware.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 550-564"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817526","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Multi-channel enhanced graph convolutional network for sentiment analysis on instrumental music descriptions","authors":"Fangge Lv , Huasang Wang","doi":"10.1016/j.aej.2025.03.088","DOIUrl":"10.1016/j.aej.2025.03.088","url":null,"abstract":"<div><div>Traditional single-channel feature extraction methods face challenges in instrumental music sentiment analysis, primarily due to their reliance on a single type of dependency, which overlooks the complex relationships between musical elements and emotions. While graph convolutional network (GCN)-based approaches show potential, they still struggle with aggregating both musical structure information and emotional details, especially in instrumental music without lyrics, where misinterpretation of emotional features is common. Moreover, insufficient domain knowledge hinders the model’s ability to capture subtle differences in musical terminology, further reducing sentiment analysis accuracy. To address these challenges, we propose a sentiment analysis graph neural network called KSD-GCN, where the sentiment-enhanced syntactic graph convolution module enriches the dependency graph by integrating external sentiment knowledge, thereby improving the model’s ability to capture emotions. The dependency relation embedding module focuses on capturing syntactic dependency information within the sentence. Additionally, we introduce a multi-layer interactive attention mechanism that effectively integrates syntactic, dependency, and semantic information. Through this interaction, the model can finely capture the sentiment and syntactic structure of the sentence at different layers, significantly improving the accuracy of aspect-based sentiment analysis. Experimental results show that, on multiple datasets, the model outperforms baseline models across several performance metrics.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 527-539"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817520","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Optimizing crystallization processes in desulfurized gypsum using zinc oxide for superior construction gypsum","authors":"Guihai Gao, Jianxu Chen, Bo Qian","doi":"10.1016/j.aej.2025.04.001","DOIUrl":"10.1016/j.aej.2025.04.001","url":null,"abstract":"<div><div>This study investigates the relationship between crystal characteristics and performance enhancement in construction gypsum through zinc oxide modification of desulfurized gypsum. Systematic experiments demonstrated that ZnO addition effectively controlled crystal formation and improved material properties across multiple parameters. The incorporation of 0.6 % ZnO facilitated complete conversion to α-hemihydrate at temperatures between 160 and 200 °C, with optimal crystallization observed at 180 °C. Microstructural analysis revealed that ZnO modification promoted the formation of interlocking needle-like crystals with lengths of 20–30 μm and widths of 1–2 μm after 7 days of hydration. The modified gypsum exhibited significantly enhanced performance characteristics, including reduced water demand (66.8 % vs 71.2 % for control), moderate setting times (initial 8.2 min, final 13.5 min), and superior mechanical properties. The progressive strength development showed continuous improvement over time, with 7-day flexural and compressive strengths reaching 4.8 MPa and 21.5 MPa, respectively. The modified crystal structure demonstrated improved packing density and reduced porosity (35.2 % compared to 38.5 % in unmodified samples), contributing to enhanced material performance. These findings provide valuable insights for industrial-scale production of high-performance construction materials from desulfurization byproducts.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 484-493"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817522","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abdulrhman M. Alshareef , Aisha Alsobhi , Alaa O. Khadidos , Khaled H. Alyoubi , Adil O. Khadidos , Mahmoud Ragab
{"title":"Automated detection of ChatGPT-generated text vs. human text using gannet-optimized deep learning","authors":"Abdulrhman M. Alshareef , Aisha Alsobhi , Alaa O. Khadidos , Khaled H. Alyoubi , Adil O. Khadidos , Mahmoud Ragab","doi":"10.1016/j.aej.2025.03.139","DOIUrl":"10.1016/j.aej.2025.03.139","url":null,"abstract":"<div><div>In the digital era, differentiating text produced by Chat Generative Pre-Trained Transformer (ChatGPT) from human-produced text is critical in a digital setting. As artificial intelligence (AI) increasingly produces content, discriminating between sources becomes significant to prevent spam, improve data accuracy, control content quality, and ensure data reliability. Deep learning (DL), machine learning (ML), and Natural Language Processing (NPL) approaches can distinguish between AI and human-generated text based on superior linguistic context, signals, or patterns frequently used. The ability to proficiently make this alteration has huge achievement effects, from enhancing user contribution to contrasting disinformation and upholding the reliability of online communication platforms. This research paper presents a new Gannet Optimization Algorithm with DL-based detection and classification (GOA-DLDC) technique for ChatGPT and human-generated text. The main objective of the GOA-DLDC technique is to recognize and classify the human and ChatGPT-generated text. The GOA-DLDC technique employs the BERT approach for feature vector generation. The classification method is also implemented using the convolutional gated recurrent unit (CGRU) model. To enhance the classification performance of the CGRU model, the hyperparameter-tuning procedure is executed using the gannet optimization algorithm (GOA). The experimental validation of the GOA-DLDC methodology is performed on a dataset comprising human and ChatGPT-generated text. The investigational outcome of the GOA-DLDC methodology portrayed a superior accuracy value of 94.90 % and 94.40 % under human and ChatGPT datasets.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 495-512"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817523","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Marya Zainab , Adnan Aslam , Takasar Hussain , Muhammad Ozair , Ahmed M. Shehata , Kottakkaran Sooppy Nisar , M. Abdalla
{"title":"A fractional-operator approach for unraveling rabies disease dynamics in animal population","authors":"Marya Zainab , Adnan Aslam , Takasar Hussain , Muhammad Ozair , Ahmed M. Shehata , Kottakkaran Sooppy Nisar , M. Abdalla","doi":"10.1016/j.aej.2025.03.082","DOIUrl":"10.1016/j.aej.2025.03.082","url":null,"abstract":"<div><div>Rabies remains a significant public health threat. It necessitates robust mathematical frameworks to understand its transmission dynamics and control strategies. In this work. we develop a novel fractional order mathematical model for rabies disease considering Atangana–Baleanu fractional operator to better capture the memory and hereditary properties of the disease spread. The existence and uniqueness of solutions are established by applying rigorous analytical techniques, ensuring the model’s well-posedness. Additionally, a numerical scheme developed by Toufik and Atangana has been applied to solve the fractional order system efficiently. Numerical simulations illustrates the impact of fractional dynamics on disease propagation providing deeper insights into rabies control mechanisms. The results highlights the advantage of fractional modeling in capturing complex epidemiological behaviors compared to classical integral order models. The findings contribute to the ongoing efforts in developing more accurate predictive measures for the Rabies control and prevention.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 540-549"},"PeriodicalIF":6.2,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143817525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Automatic motion recognition technology based on fuzzy clustering algorithm and VR video image","authors":"Ganbin Xu , Jianwei Lin","doi":"10.1016/j.aej.2025.03.083","DOIUrl":"10.1016/j.aej.2025.03.083","url":null,"abstract":"<div><div>The rapid development of Virtual Reality (VR) technology has heightened interest in immersive video applications. Current dynamic video segmentation methods emphasize iterative pixel-level classification based on fuzzy membership levels. Traditional motion recognition, however, struggles with posture variations in complex scenarios, leading to inaccuracies. To address this, we propose a fuzzy clustering-based framework for VR video motion recognition, enhancing accuracy through optimized pixel segmentation and preprocessing. The system’s stability across varying frame rates was evaluated. Experiments utilized a CCD-captured database of six motion poses (sitting, squatting, kneeling, etc.). Results demonstrated 96% accuracy for squat detection, 6.7% false positives for kneeling postures, and a 2%–8% error range for sitting poses—significantly higher than other categories due to postural variability. This study advances VR motion recognition by integrating fuzzy clustering with adaptive image processing, offering robust solutions for immersive training, rehabilitation monitoring, and human–computer interaction systems.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"124 ","pages":"Pages 462-469"},"PeriodicalIF":6.2,"publicationDate":"2025-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143792135","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}