ptimized Insider Threat Classification and Secure Cookie File Transfer Using SHA-3 Merkle Tree and CNN-LSTM Hybrid Models

Abstract

Background Information : This is the most significant threat to the security of organizational 
data, namely insiders. The latest techniques are becoming more and more challenged in terms 
of precision and scalability. A hybrid architecture that combines SHA-3 Merkle Tree and CNN
LSTM offers a resilient solution for the secure transmission of the cookie file and insider threat 
classification. 
Objectives: The framework aims to improve the accuracy of insider threat detection, guarantee 
secure data communication, minimize false positives and negatives, and provide real-time 
scalability in applications in dynamic environments. 
Methods: The system uses SHA-3 Merkle Tree for data integrity verification and the CNN
LSTM hybrid model with an attention mechanism for the classification of insider threats, with 
efficiency and robustness guaranteed. 
Empirical Results : Accuracy of 97.8 % Slightly at the price of a relatively quite small false 
positive and false negative (2.7% and 2.3% respectively); with great scaling performance, 
around 92.5%. 
Conclusion: Hybrid architecture happens with great security and efficiency towards the 
detection of insider threats as well as the secure file transfer, thus fitting well in the real time 
applications. Future scope can be seen through adaptive learning mechanisms and quantum 
resistant cryptography mechanisms.