AI-Powered Anomaly Detection in Oracle Database: Leveraging Machine Learning for Proactive Threat Mitigation

Abstract

Lately, the vast increase in data and the rising complexity of business systems have made traditional security systems in database management less practical. Because Oracle Database is widely employed in businesses, it is commonplace where valuable and critical business data is found. Any unusual activity from persons within the organisation, bad misconfigurations or outsiders can cause serious problems. It introduces a reliable system for detecting abnormal activities in Oracle Databases with ML. This system uses past access records, SQL logs and system usage data to identify situations that need further examination. A set of approaches that fuse supervised learning with unsupervised learning using Isolation Forests, Autoencoders and LSTM systems is presented in this article. Based on comparative studies, our method produces dependable results with a small number of mistakes. In addition, we suggest a way to set up AVDF that easily connects to both Oracle Audit Vault and Database Firewall (AVDF). The research demonstrates that AI plays an important role in making databases safer and highlights what improvements could be made in the future.v