Research On Security Auditing Technology Of Database Based On Oracle

Database is the core and basis of information systems, which stores important information most organizations live on. The importance of protecting database secuirty is now drawing more and more attention. At present, most information systems have taken certain safety protection measures to ensure the security of database. But only safety protection is not enough. When attack happens, we have, at least, to know how the system is attacked and how to restore the database. Besides, what specific leak exists in the system, how to make the system alert to the attacks and how to obtain the evidence of the attackers have to be awared. Thus, database security auditing technology emerges as required.This paper discusses the satety standard of database and the requirement this standard has put forward to the security auditing; moreover, it studies the safety mechanisms of Oracle, especially the security auditing mechanisms. Oracle can record all the operations within the system and offer complete auditing data. However, when handling a mass of auditing data, Oracle can't provide analysis tools to analyse those auditing data and abstract the useful information from the many data.In order to improve the existing security auditing mechanisms of Oracle, the paper founds a model of users'natural behavior to detect users'action. After studying the technology of data mining and connect the specialty of audit records of oracle database, this paper advances a way bases on association rules and sequence pattern to found the users'natural behavior model. In this paper, in order to make the arithmetic is suitable to analysis the audit data, it alters the key steps of Apriori, AprioriAll, GSP and PrefixSpan. At last, the paper designs an Oracle auditing system, it can analysis the audit data and found users'natural behavior to detect users'action.