Blockchain-Enabled Smart Certificate Authentication System with Performance Prediction using DLSTM

Abstract

The integration of secure credential management and intelligent analytics is vital for modern educational ecosystems. Blockchain ensures immutable, decentralized, and tamper-resistant storage of student records while preserving privacy through cryptographic techniques, enabling secure verification without disclosing sensitive information. On the analytics front, DLSTM models are trained on temporally structured academic datasets to capture sequential patterns influencing student success. The experimental evaluation compares DLSTM with conventional machine learning models including Random Forest and Support Vector Machine (SVM). The results indicate that DLSTM achieves superior predictive performance with an accuracy of 97.7%, macro precision of 95.13%, macro recall of 96.75%, and macro F1-score of 95.89%. In comparison, Random Forest and SVM achieved accuracies of 96.2% and 93.6%, respectively, with lower recall and F1-scores.