采用填充字符的频繁序列模式挖掘算法  

Frequent Sequence Pattern Mining Algorithm Adopting Dummy Characters

作　　者：张君雁[1,2] 闵帆[3] 

机构地区：[1]成都大学信息科学与技术学院,四川成都610106 [2]模式识别与智能信息处理四川省高校重点实验室,四川成都610106 [3]电子科技大学计算机科学与工程学院,四川成都610054

出　　处：《成都大学学报（自然科学版）》2013年第2期134-137,共4页Journal of Chengdu University（Natural Science Edition）

摘　　要：具有固定通配符间隔的频繁序列模式挖掘算法应具有Apriori属性,从而保证在实际应用中能挖掘出有意义的长模式.而原有的问题定义集合存在一定的不足阻碍了该属性的实现.通过引入填充字符改变部分问题定义,解决原定义引起的一些极端性问题,并在模式挖掘过程中保证了完整性和有效性.将基于新定义集合提出的FSPM算法与基于原定义集合的MMP算法分别在DNA序列上进行实验,结果表明算法实现了Apriori属性.It is necessary for frequent sequence pattern mining algorithm which has fixed wildcard interval to implement Apriori property, which makes sure that the long and interesting patterns can be mined in practical applications. The drawback of the original problem defmition set hinders the realization of Apriori property. In this paper, some dummy characters are employed to modify part of the definitions. Some ex- treme problems caused by the original definitions are solved. The integrity and effectiveness are ensured in the process of frequent sequence mining. The comparison experiments are executed between FSPM algorithm based on new definition set and MMP algorithm based on original definition set in DNA sequences. The re- sults show that FSPM algorithm realizes Apriori property.

关 键 词：序列模式挖掘 填充字符 固定间隔 Apriori属性 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]