基于航站楼门禁日志挖掘的物理入侵检测技术  

PHYSICAL INTRUSION DETECTION TECHNOLOGY BASED ON TERMINAL BUILDINGS ACCESS LOG MINING

作　　者：顾兆军[1] 安一然 刘飞[1] 

机构地区：[1]中国民航大学计算机科学与技术学院,天津300300

出　　处：《计算机应用与软件》2015年第11期317-320,324,共5页Computer Applications and Software

基　　金：中国民航局科技项目(MHRD201128);中国民航大学科技项目(3122013C004);中国民航大学科研启动项目(2013QD24X);中央高校基本科研业务费中国民航大学专项资助项目(3122013C004)

摘　　要：针对国内大型航站楼门禁系统的安全性需求,以及门禁日志数据的特殊性,传统的数据挖掘方法难以取得良好的效果,为解决这一问题,提出一种改进的适用于门禁日志数据的Prefix Span算法。实验结果表明,此算法可以很好的建立正常行为模式,然后运用基于先序遍历树的子序列精确匹配定位算法检测异常行为模式,实现了异常入侵检测,显著改善航站楼的安全防范能力及机场的工作效率。For the security requirements in domestic large-scale terminal buildings＇ access control system, and for the particularity of access control log data, traditional data mining methods are difficult to obtain good results. To solve this problem, we bring forward an improved PrefixSpan algorithm, ~hieh is applicable to the log data. Experimental results show that the algorithm can be very good in setting up normal behaviour patterns. Then we make use of subsequence exact matching positioning algorithm based on pre-order tree traversal to detect abnormal behaviour patterns, and realise the abnormal intrusion detection, this significantly improves the security capacities of terminal buildings and the working efficiency of airports.

关 键 词：机场航站楼 门禁管理系统 PREFIXSPAN算法 物理入侵检测 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]