改进的最大频繁项目序列集挖掘算法

Improvement of ISS-DM Algorithm based on Mining of Maximal Frequent Item Sequence Sets

影响关联规则挖掘的关键问题是最大频繁项目序列集的生成问题,而传统的算法往往要求对事务数据库进行多次扫描,从而提高了I/O代价.阐述了项目序列集和它的基本操作的定义,然后详细描述了ISS-DM的最大频繁项目序列集生成算法,并在此基础上提出了一种改进的ISS-DM算法,最后进行了相应的验证.实践证明,改进后的算法同原算法相比,对相同的数据量进行挖掘,算法执行时间明显减少,效率较高.

Impact of mining association rules is the key to the production problems of maximal frequent item sequence sets, and the traditional algorithms often require a number of scanning database services, thereby improving of I/O costs. This paper described the item sequence sets and the operation of its basic definition, and detailed the ISS-DM algorithm based on mining of maximal frequent item sequence sets： An improved algorithm is proposed for ISS-DM, and the corresponding certification with the same amount of data processing of original algorithm, the improved algorithm has significantly reduced of algorithm implementation time.