面向异常检测的关联模式快速挖掘算法

ANOMALY DETECTION-ORIENTED RAPID MINING ALGORITHM OF ASSOCIATION PATTERN

随着网络和其它信息技术的广泛应用,网络数据流量急剧增长,但现有网络流量异常监测的准确性与实时性均达不到实际应用的需求,迫切需要对流量数据进行快速、深层次的分析。因此,提出一种快速关联模式挖掘算法,通过提取重要的网络数据特征进行关联挖掘,不仅为流量数据分析判断提供及时准确的参考和借鉴,而且提高了监测准确性和效率。

With the rapid growth of network and other information technology, net-flow increases greatly. The existing net-flow anomaly detection cannot satisfy the practical application requirements in both accuracy and reality, a quicker and deeper analysis for net - flow data becomes urgent. This paper proposed a rapid association pattern mining algorithm. By the associated mining based on selected important network data properties, the algorithm offers timely and precise references to the analysis and judgement on net-flow data, and improves the accuracy and efficiency of the detection.