A blockchain-based deep learning approach for student course recommendation and secure digital certification

Over the past decade, the student course recommendation process with secure certificate issuance has remained a critical research area due to the rise of e-learning and personalized learning. The recommendation system enhances the recommended educational resources to improve the students' learning process. The previous conventional research works shared hybrid content and collaborative filtering techniques, which boosted academic performance, personalized learning, and secure certification for students. However, the existing techniques faced several difficulties in handling the syllabus updates based on evolving recommendations, complexity, and security issues related to certificate issuance. To address the challenges in the existing techniques, the research introduces the Deep Certifier-DX509 model for secure certificate issuance and student course recommendation. The proposed approach exploits the Modified Attention-Enabled Deep Long Short-Term Memory (MA-DLSTM) Model as a recommendation system to suggest the most suitable courses based on users' prior academic performance, and integrates X509 as the Certificate generation algorithm. Specifically, the incorporation of the X509 Blockchain with Proof-of-Work (PoW) in the certificate sub-system serves as a major contribution to enhance the security with Two-step authentication and generates accurate course recommendations. Experimental results demonstrate that the proposed Deep Certifier-DX509 model shows superior performance, achieving a high Genuine User Rate (GUR) of 0.73, Memory Usage of 453.81KB, Transaction time of 1.03 s, Responsiveness of 2.39s and Throughput of 119.52bps, outperforming the other existing techniques.