E ComShield: Enhancing E Commerce Cybersecurity through Cloud Native Threat Detection Frameworks

The rapid expansion of digital commerce platforms has led to a corresponding rise in cyber threats targeting e-commerce infrastructures, both in scale and sophistication. To address these challenges, this paper introduces E-ComShield, a cloud-native security framework designed to continuously observe, classify, and mitigate complex cyberattacks in online retail environments. The proposed system incorporates an Adaptive Hybrid Deep Learning model with Federated Feature Fusion (AHDL-FFF), which integrates Convolutional Neural Networks, Long Short-Term Memory networks, and Deep Neural Networks to capture spatial correlations, temporal dependencies, and high-level abstractions from diverse transactional data streams. To enhance learning effectiveness and detection reliability, E-ComShield employs advanced data preparation techniques, including missing value handling, Z-score-based anomaly filtering, categorical label transformation, and Min–Max normalization. A federated learning strategy is adopted to enable collaborative model training across distributed data sources while preserving data privacy and maintaining compliance with confidentiality requirements. Extensive experimental evaluation demonstrates that the proposed framework significantly outperforms conventional approaches, achieving an overall accuracy of 99.65%, with consistently higher precision, recall, and F1-score compared to baseline models such as Support Vector Machines, Naïve Bayes classifiers, and standalone CNN architectures. Further analysis using heat-map visualization and feature-importance ranking confirms the model’s ability to effectively differentiate among diverse threat categories, including distributed denial-of-service attacks, session hijacking attempts, and automated bot intrusions, with minimal misclassification. The deployment of E-ComShield within a cloud-native CI/CD pipeline enables scalable operation, real-time adaptability, and seamless integration of evolving threat intelligence. The results validate E-ComShield as a robust and scalable defense mechanism for modern e-commerce systems, while also paving the way for future research into autonomous, self-healing, and fully federated cybersecurity solutions for digital commerce ecosystems.