Research On Intrusion Detection Technology Based On Database

With the fast development of computer and network, the application of database is full abroad; the database get down to various region, more and more enterprises and scientific research institutes join their databases to the Internet. However the security problem of data came along with this. The security problem of data in different kinds of application system's database, the problem of anti steal and anti distort are taken highly attention by people. Conventional firewall technology can't satisfaction the security need of network. The application level intrusion aimed at application and background database is becoming more and more rampancy, such as SQL injects, script attach across site and unwarranted user access. All this intrusion could attack data source bypass foreground security system. We badly need of research about intrusion detection technology aimed at database in order to improve the security of database.In this paper, some knowledge of intrusion detection, data mining, database security, intrusion tolerance are introduced. The database application security technology is posed aimed at the present threat of network. This security technology apply conventional intrusion detection system notion to database (namely application). The core of this paper is database intrusion detection system, this system use knowledge of data mining to compare user history behavioral model with user present behavioral model, use the similarity degree to detect the present behavioral model is abnormity or not, use intrusion tolerance technology proceed protect when the intrusion is inevitability, make the system continuing provide important serve when confront with attach and destroy. This system realizes non-supervision learning through clustering mining.