This research explores the determinants affecting academic researchers’ acceptance of AI writing tools using the Theory of Reasoned Action (TRA). The impact of attitudes, subjective norms, and perceived barriers on researchers’ intentions to adopt these technologies is examined through a cross-sectional survey of 150 researchers. Structural Equation Modeling (SEM) is employed to evaluate the measurement and structural models. Findings confirm the positive influence of favorable attitudes and subjective norms on intentions to use AI writing tools. Interestingly, perceived barriers did not significantly impact attitudes or intentions, suggesting that in the academic context, potential benefits may outweigh perceived obstacles to AI writing tool adoption. Contrarily, perceived barriers do not significantly affect attitudes and intentions directly. The TRA model demonstrates considerable explanatory and predictive capabilities, indicating its effectiveness in understanding AI writing tool adoption among researchers. The study’s diverse sample across various disciplines and career stages provides insights that may be generalizable to similar academic contexts, though further research with larger samples is needed to confirm broader applicability. Results offer practical guidance for tool developers, academic institutions, and publishers aiming to foster responsible and efficient AI writing tool use in academia. Findings suggest strategies such as demonstrating clear productivity gains, establishing AI Writing Tool programs, and developing comprehensive training initiatives could promote responsible adoption. Strategies focusing on cultivating positive attitudes, leveraging social influence, and addressing perceived barriers could be particularly effective in promoting adoption. This pioneering study investigates researchers’ acceptance of AI writing tools using a technology acceptance model, contributing to the understanding of technology adoption in professional contexts and highlighting the importance of field-specific factors in examining adoption intentions and behaviors.
The proliferation of Internet of Things (IoT) devices and fog computing architectures has introduced major security and cyber threats. Intrusion detection systems have become effective in monitoring network traffic and activities to identify anomalies that are indicative of attacks. However, constraints such as limited computing resources at fog nodes render conventional intrusion detection techniques impractical. This paper proposes a novel framework that integrates stacked autoencoders, CatBoost, and an optimised transformer-CNN-LSTM ensemble tailored for intrusion detection in fog and IoT networks. Autoencoders extract robust features from high-dimensional traffic data while reducing the dimensionality of the efficiency at fog nodes. CatBoost refines features through predictive selection. The ensemble model combines self-attention, convolutions, and recurrence for comprehensive traffic analysis in the cloud. Evaluations of the NSL-KDD, UNSW-NB15, and AWID benchmarks demonstrate an accuracy of over 99% in detecting threats across traditional, hybrid enterprises and wireless environments. Integrated edge preprocessing and cloud-based ensemble learning pipelines enable efficient and accurate anomaly detection. The results highlight the viability of securing real-world fog and the IoT infrastructure against continuously evolving cyber-attacks.
The proliferation of Internet of Medical Things (IoMT) devices has created cybersecurity challenges that requiring advanced threat detection techniques along with preserving patient privacy. This paper introduces FedMedSecure, a federated few-shot learning framework to provide privacy-preserving and collaborative learning, explainable AI, and adaptive ensemble mechanisms for IoMT cybersecurity. Our approach combines CrossTransformer with learnable attack signature queries, FEAT, RelationNetwork with adaptive prototypes, and regularized MAML within a confidence-weighted ensemble architecture. The framework implements differential privacy with \((\epsilon , \delta ) = (1.0, 10^{-5})\) while achieving 75% communication reduction through efficient gradient compression. The evaluation implemented on two datasets-CICIoMT2024 (8.7M healthcare IoT samples across 19 attack categories) and CIDC2017 (2.8M general IoT samples across 14 attack categories)-We have achieved an exceptional performance as the following: 99.9% accuracy on CICIoMT2024 and 93.3% on CIDC2017 in supervised learning, 99.7-99.8% and 91.0-99.3% respectively in few-shot scenarios, and 99.8% while the global accuracy in federated learning experiments across 8 institutions. Cross-dataset validation confirms robust generalization capabilities, with few-shot learning achieving rapid adaptation from 91.0% with 5 shots to 99.3% with 50 shots on CIDC2017. Counterintuitively, the original 19-class taxonomy outperformed theoretically optimized 5-class clustering in few-shot learning, providing new insights for meta-learning research. The multi-level explainable ai (XAI) framework shown the packet timing and protocol features as primary discriminators, and shown analyst trust. Our FedMedSecure enables collaborative healthcare cybersecurity without compromising privacy that establishing a new paradigm for trustworthy AI in sensitive domains like healthcare with broader applicability to financial services, critical infrastructure, and government networks that requiring privacy-preserving collaborative threat detection.
Phishing attacks continue to evolve rapidly, with new campaigns emerging faster than traditional detection systems can adapt. Existing machine learning approaches require extensive labeled datasets, creating vulnerability windows when novel attack patterns appear. This limitation is particularly problematic in cybersecurity where obtaining large labeled datasets for emerging threats is time-consuming and expensive. This paper presents XF-PhishBERT, an explainable few-shot learning framework for phishing detection that combines ModernBERT transformer architecture with domain-specific URL features. The approach integrates prototypical networks and model-agnostic meta-learning (MAML) to enable effective detection with minimal training examples. A consensus-based feature selection methodology combines Random Forest importance, Mutual Information, and Recursive Feature Elimination with Cross-Validation (RFECV) to identify optimal feature subsets. The framework incorporates comprehensive explainability through SHAP analysis, attention visualization, and counterfactual explanations. Experimental evaluation on two datasets demonstrates that XF-PhishBERT achieves 99.9% accuracy with 10 examples per class and maintains 98.5% accuracy in one-shot learning scenarios. Cross-dataset evaluation shows 186% performance retention compared to 39% for traditional methods. Ablation studies confirm the contribution of each component, with ModernBERT integration providing 4.54 percentage point improvement over baseline approaches. Real-world deployment through a browser extension validated practical utility with 98.3% precision and 42ms average latency. The results demonstrate that few-shot learning can address the fundamental challenge of limited labeled data in cybersecurity applications while providing transparent decision-making support for security analysts.
Respiratory disease diagnosis remains challenging in resource-constrained settings, where limited specialist expertise contributes to diagnostic uncertainties affecting over 300 million people worldwide. This study presents E-RespiNet, a novel multi-modal deep learning architecture that integrates ELECTRA’s discriminative pre-training with a triple-stream convolutional neural network framework, enhanced through Harmony Search with Opposition-Based Learning optimization for automated respiratory sound classification. The architecture simultaneously processes mel-frequency cepstral coefficients, discrete wavelet transforms, and mel-spectrograms through parallel CNN streams, with features integrated through hierarchical fusion and ELECTRA-based contextual enhancement. Comprehensive evaluations on two independent clinical datasets—the Asthma Detection Dataset Version 2 (1,211 recordings across five conditions) and King Abdullah University Hospital dataset (940 samples from 81 subjects across four conditions)—demonstrated exceptional performance with 98.9% and 98.8% accuracy respectively, representing 5.0% and 4.3% improvements over baseline configurations. Cross-institutional validation revealed 75.7% average accuracy with a 23.3% generalization gap, substantially better than typical medical AI cross-domain performance. Gradient-weighted class activation mapping provided clinically relevant interpretability, while the Harmony Search optimization framework enhanced both performance and cross-institutional robustness. These results demonstrate significant advances in automated respiratory sound analysis through discriminative language model integration and metaheuristic optimization, establishing important benchmarks for deployable respiratory diagnostic tools in diverse healthcare settings.