一种有效的序列模式增量式更新方法

Efficient incremental updating method for sequence pattern

针对序列模式增量式更新挖掘算法产生大量候选项集以及多次扫描数据库的问题,提出了一种有效的增量式更新算法ESPIA,该算法利用基于2-序列矩阵挖掘算法ESPE对原数据库和增加数据库一次扫描产生序列模式,通过对频繁模式和非频繁模式进行相应的剪枝减少了序列的比较和扫描次数,降低了算法时间和空间复杂度,实验证明该算法是有效和准确的。

For sequential pattern mining algorithm for incremental updating designates a large amount of options set and repeatedly scans the database,this paper proposes an Efficient Sequence Pattern Incremental updating Algorithm（ESPIA）.This algorithm uses ESPE which is based on the 2-sequence matrix to generate sequence patterns by scanning the original database and the increase database only once.Then through pruning the frequent and non-frequent patterns it reduces the times of comparison and scanning the sequence,so it lowers the time and space complexity.Experiments show that this algorithm is effective and accurate.