Based on the user behavior characteristics of mining database anomaly detection model design

This paper first designs a database anomaly detection model. This model can be more accurate depiction of the behavior of users and can improve database abnormal detection correctness. In the design of the model, the Apriori-kl algorithm is put forward and it is based on the clustering k-means algorithm and the improved Apriori algorithm. It can better exploit the behavior of users, and the database abnormal more effectively detection. Through the related experiments show that Apriori-kl algorithm in time expenses and detection accuracy in is better than pure use based on association rules mining technology Apriori algorithm.