Secure Framework for Cloud based E-Education using Deep Neural Networks

E-Education is an important aspect that facilitates the online and virtual education. The advancements in the education institutes are increasing tremendously which requires high-end servers, software's, and applications that leads to the huge investment cost. The optimized solution is adoption of cloud services. Hosting the applications and documents over the cloud provides flexibility, performance, and quality of the education. Despite the advantages, the main challenge associated with the E-Education over the cloud is privacy and security of the outsourced data. In the existed solutions stores all the documents in the same storage area, similar type of authentication algorithm for all users, and single type of encryption algorithm is applied for all documents. In this paper, a secure framework of E-Education over the cloud environment is proposed that performs the authentication of the cloud users in the effective way through the multiple-factors based on the role and activities of a user. To extract the accurate features of the biometric face modality VGG2F Convolutional neural network is applied. Documents storage classification, secure question paper generation, and verification are proposed in the secure way. The proposed model reduces the overhead, processing time, and billing cost by adopting different deployment models, authentication mechanisms, and cryptographic ciphers. The biometric face authentication through the VGG2F model improves the accuracy, and hybrid combination of encryption algorithms with different key sizes applied on the documents resists various attacks. The evaluated results show that the proposed framework is suitable for E-education in the cloud.