An Attention-Driven Hybrid Deep Learning Framework for Intrusion Detection in Cloud-Based E-Learning

Abstract

E-learning environments on clouds are scalable and flexible, yet are prone to cyberattacks. The current research suggests an Attention-based CNN-BiLSTM Hybrid Intrusion Detection System (ACH-IDS) for the protection of e-learning environment based on clouds from cyberattacks. The system consists of a CNN component for feature extraction from space dimensions, a BiLSTM layer for capturing features from time dimensions, and an Attention component for learning the attacks better. Metrics such as ROC-AUC, F1-score, accuracy, precision, and recall are used to evaluate the proposed model's performance after data pre-processing. There was a ROC-AUC of 99.81%, a precision of 99.21%, a recall of 99.35%, an F1-score of 99.28%, and an accuracy of $\mathbf{9 9. 4 2} \boldsymbol{\%}$ found in the testing.