Multi Relational Mining in Network Intrusion Detection

The network intrusion detection (NIDS) is faced with the question to detect many kinds of intrusion. In order to detect the complex attack, network intrusion detection system need to analysis massive data captured form different network safety equipments. So a new multi relational mining algorithm MRA2 is proposed. MRA2 depend on the association rules mining technology and the probability function dependency method which is proposed through extending the theory of function dependency. MRA2 is able to synthesize the various datalog resources to detect intrusion effectively and reappear to the complex network attack scenario.