基于数据库用户行为分析的改进FP-Growth算法

数据库是信息系统的核心，是最吸引攻击者的目标，其用户行为记录是一种特定的类型，有相对固定的成份。FP-Growth算法在规则挖掘时会产生一些冗余的、无意义的规则。本文首先给出数据库用户行为的定义，将数据库的用户行为属性按重要性阈值排序，并从中选取关键属性或属性组，在FP-Growth算法的基础上提出一种基于用户行为分析的BFP-Growth算法，避免产生无意义的规则，节省了存储空间和时间，提高了挖掘效率。

An improved FP-Growth algorithm based on Database user behavior analysis

The Database is the core of the information systems and it is the most attractive target of the attackers. Its user behavior records is a kind of specific types,has the relatively fixed ingredients.FP-Growth algorithm will produce some redundant and meaningless rules when it excavates rules.This article sorts Database user behavior properties according importance threshold of properties.It selects key attribute or attribute group from the properties.This paper presents the definition of Database user behavior,and brings up an improved FP-Growth algorithm based on user behavior analysis named BFP-Growth algorithm. The new algorithm avoid to produce meaningless rules, saves storage space and time and improve the mining efficiency.