statistical view, and the evaluations of discovered knowledge
are also enhanced, so as to remove the meaningless
knowledge, add the prior knowledge, and make the
discovered knowledge more useful.
d) Extraction of Security Situation Correlation Rules
The knowledge obtained from the process of knowledge
discovery is transformed to the correlation rules of security
situation by adding the correlation actions, and can be
applied to the online correlation analysis of network
security situation. Firstly, the strong correlation rules
among the attributes of alert events are analyzed, which are
extracted by FP Tree algorithm. If this kind of rules is
related to some regular access, then the deletion action is
added, and these rules are transformed to the filtering rules
of alert events. Secondly, the sequential relationships
among alert events are analyzed, which are extracted by
WINEPI algorithm. If this kind of sequential relationship is
associated with some type attack and the combination rules
of attack events is formed, then the new security attack
event is added. Finally, the generated correlation rules are
transformed to the formal rule encodings, and added to the
online correlation knowledge base.
statistical view, and the evaluations of discovered knowledge
are also enhanced, so as to remove the meaningless
knowledge, add the prior knowledge, and make the
discovered knowledge more useful.
d) Extraction of Security Situation Correlation Rules
The knowledge obtained from the process of knowledge
discovery is transformed to the correlation rules of security
situation by adding the correlation actions, and can be
applied to the online correlation analysis of network
security situation. Firstly, the strong correlation rules
among the attributes of alert events are analyzed, which are
extracted by FP Tree algorithm. If this kind of rules is
related to some regular access, then the deletion action is
added, and these rules are transformed to the filtering rules
of alert events. Secondly, the sequential relationships
among alert events are analyzed, which are extracted by
WINEPI algorithm. If this kind of sequential relationship is
associated with some type attack and the combination rules
of attack events is formed, then the new security attack
event is added. Finally, the generated correlation rules are
transformed to the formal rule encodings, and added to the
online correlation knowledge base.
การแปล กรุณารอสักครู่..
