A new approach for intrusion detection

With the wide spread of Internet, the attacks against computers and network infrastructures have become an increasingly serious problem. Aiming at the detection technology, we present a detection algorithm STSW based on a time-sensitive sliding window, which expands the application of data mining in intrusion detection. Using KDD CUP99 as experimental data, the paper studied the effect of selection of different parameters on detection results. The efficiency of the presented algorithm was compared with that of the algorithm based on SPADE mining sequence patterns. Experimental results show that the presented intrusion detection algorithm can get quite satisfactory effect. The performance efficiency of the algorithm in this paper is better than that of the algorithm based on SPADE.