User behaviors attributes of database anomaly detection model

摘要

This paper includes the description of designing a data-base anomalydetection system, whichis capable of being more precisein depictingthe behaviors of individuals and improving data-base abnormaldetecting correctness. In designing the system, the Aprioriapproach is used first and depends on the k-meansclusteringand the Apriorimethods.It is capable of more efficiently exploitingusers'behaviors, and the data-base abnormalmore efficient detecting. The relevantstudiesshow that Apriorimethodaccording to time efficiency and precision of detectionis more optimal than thesoleutilizationaccording to association rulesmining approachesApriorimethod.