一种基于规则表达式约束的序列模式增量式挖掘算法

An incremental mining algorithm for sequential pattern based on regular expression constraints

序列模式挖掘是数据挖掘中的研究热点之一。在挖掘过程中需要用户的参与日益显得重要。为了提高挖掘过程中的交互性,本文提出了一个基于规则表达式约束的序列模式增量式挖掘算法RE_IncUp。该算法首先利用约束对已经挖掘出的频繁序列模式进行预处理,缩小了搜索范围;然后采用模式扩展方法把规则表达式约束和增量挖掘过程融为一体,并且采用先修剪后计算支持度的方法进一步缩小了搜索范围,降低了支持度的计算量。该算法允许用户不断改变约束条件,实现交互式挖掘而且可将挖掘的目标仅仅聚焦到用户感兴趣的模式上。实验表明该算法对序列模式的维护和满足用户的需求都是十分有效的。

Sequential pattern mining is an important and active issue in data mining. The user＇s participation in the procession of mining is more and more important. In order to reduce response time, an algorithm, RE_IncUp, is proposed in this paper. Firstly, to reduce the searching space, this constraint is used to pre-process the frequent sequential patterns mined previously. Secondly, to decrease the amount of counting, the first-pruning and then-counting method is adopted to reduce the searching space enormously. The method allows users to change constraints to implement interactive mining, and also facilitates the users to focus the mining on their interesting patterns. The experimental analysis shows that the new algorithm is more efficient in the maintenance of the sequential pattern and requirement of users.