一种序列模式增量式挖掘算法  

An Incremental Sequence Pattern Mining Algorithm

作　　者：付仲良[1,2] 陈楠[1,3] 

机构地区：[1]武汉大学遥感信息工程学院,武汉市珞喻路129号430079 [2]武汉大学测绘遥感信息工程国家重点实验室,武汉市珞喻路129号430079 [3]湖北省环境监测中心站,武汉市八一路338号430072

出　　处：《武汉大学学报（信息科学版）》2010年第7期763-767,共5页Geomatics and Information Science of Wuhan University

摘　　要：提出了一种适用于序列数据库增长的序列模式增量式挖掘算法IU_DB。该算法将数据库更新后新产生的频繁序列分为两类分别进行挖掘,挖掘过程中充分利用前次挖掘结果来提高本次挖掘的效率。测试结果表明,IU_DB算法是正确和完备的,并且当新增数据集较小和支持度阈值较低时,IU_DB算法表现出较好的性能。Incremental sequence pattern mining algorithm is one of the hot research issues in the field of sequence pattern mining.In this paper,an incremental mining algorithm for sequence pattern named IU_DB is proposed,which deals with the growth of sequence database.IU_DB algorithm divides the newly emerged frequent sequences generated by database updating into two categories and mines them respectively.The results of previous mining are applied to improve the efficiency of the mining.Final experimental results show that the IU_DB algorithm is correct and efficient.Furthermore,when new datasets is small and threshold of support is low,IU_DB algorithm performs better.

关 键 词：数据挖掘 序列模式 增量式更新 

分 类 号：P208[天文地球—地图制图学与地理信息工程] TP311[天文地球—测绘科学与技术]