时态约束下的数据挖掘问题及算法

A Method of Data Mining Based on Temporal Constraints

对于一个大型数据库而言 ,为了提高挖掘效率 ,必须考虑减少数据库的扫描次数 ,同时使内存需求量保持在一个适当的水平上 .把时态约束应用到事务数据库的挖掘中 ,可以获得更好的效率 .本文首先利用时态区间代数操作实现原始数据库的过滤和挖掘时态区间的合并 ;然后在定义项目序列集操作的基础上 ,提出一个称为TISS DM的发现频繁项目序列集的高效算法 ;最后讨论了这个算法的效率 .

For a large database,an efficient algorithm for mining association rules should try to reduce the number of the database passes and control the size of necessary memory.Temporal constraints can help us to get higher efficiency in data mining.This paper first defines temporal interval operators,and uses them to filter the database and merge temporal intervals for mining.Then,it gives a new effective algorithm called TISS-DM for mining temporal association rules.Finally,the performance of this algorithm is discussed.