The agnostics way platform for data governance with metadata management on the big data ecosystem

This research focuses on enhancing the security of metadata in big data environments by developing a novel Recurrent Neural Data Encryption Model (RNDEM). The model addresses critical data governance and security challenges, particularly in cloud computing, where traditional encryption methods often prove inadequate. RNDEM integrates advanced techniques such as homomorphic encryption and blockchain mechanisms to protect against unauthorised access and Denial-of-Service (DoS) attacks. This model offers organisations a reliable framework for securing sensitive data in large-scale data ecosystems, supporting improved data governance and regulatory compliance. Combining neural networks with encryption algorithms in a hybrid system aims to strengthen data security while maintaining efficiency. Experimental results demonstrate significant improvements, achieving a confidentiality rate of 1.035%, a low error rate of 15.27%, an encryption time of 0.531 ms and a hash calculation time of 1.95s. Compared to existing models, RNDEM exhibits superior performance regarding privacy rate, error rate and hash computation time.