Abstract:

In view of the problems of the implementation of intrusion detection and analysis of abnormal behavior under the cloud computing internal network environment, this paper does the classification research on the cloud intrusion detection datasets (CIDD) by using Weka machine learning classification algorithms, and realizes naive Bayesian algorithm for abnormal behavior classification of internal network users through the method of software engineering. Experimental results on the classification of malicious behavior and normal behavior show that the naive Bayesian algorithm implemented in the paper achieves higher classification accuracy. The algorithm can effectively classify and analyze the internal network user behaviors of CIDD, which proves the effectiveness of the proposed scheme and algorithm.

Key words: cloud computing, user behavior, intrusion detection, machine learning, classification