| Title | Anomaly Detection in Role Administered Relational Databases — A Novel Method |
| Publication Type | Conference Paper |
| Year of Publication | 2018 |
| Authors | Ramachandran, Raji, Nidhin, R, Shogil, P P |
| Conference Name | 2018 International Conference on Advances in Computing, Communications and Informatics (ICACCI) |
| Keywords | Access Control, access control measures, anomaly detection, anomaly detection system, anomaly detection technique, authorisation, common security measures, composability, database access control, database management system, database management systems, Databases, Human Behavior, Intrusion detection, learning (artificial intelligence), Metrics, pubcrawl, Quiplet, relational database security, relational databases, Resiliency, Role administered relational databases, Role-Based Access Control database, security feature, support vector machine, Support vector machines, Training, User profile |
| Abstract | A significant amount of attempt has been lately committed for the progress of Database Management Systems (DBMS) that ensures high assertion and high security. Common security measures for database like access control measures, validation, encryption technologies, etc are not sufficient enough to secure the data from all the threats. By using an anomaly detection system, we are able to enhance the security feature of the Database management system. We are taking an assumption that the database access control is role based. In this paper, a mechanism is proposed for finding the anomaly in database by using machine learning technique such as classification. The importance of providing anomaly detection technique to a Role-Based Access Control database is that it will help for the protection against the insider attacks. The experimentation results shows that the system is able to detect intrusion effectively with high accuracy and high F1-score. |
| DOI | 10.1109/ICACCI.2018.8554752 |
| Citation Key | ramachandran_anomaly_2018 |
Anomaly Detection in Role Administered Relational Databases — A Novel Method