摘要
字符串比较是计算机信息处理的重要方法之一。针对现有关联规则挖掘算法不能记忆及利用历史挖掘成果的局限性,提出了将事务数据库转化为项目数据库,构造项目的支持事务标识符有序序列方法。为提高挖掘效率,减少串处理效率较低的负面影响,给出了双序列串比较算法,以及针对串比较的大项目频繁集发现方法。
Character string comparing is an important method in computer information managing. To solve the localization question of current algorithms for mining association rule not to save and use historical mining harvest, a method is present that transaction database is translated item database, and supporting transaction identifier sequence is constructed. An algorithm for comparing double sequence strings and a method of finding bigger frequent item set based on string comparing are given in order to improve mining efficiency and decrease negative influence that string comparing' efficiency is lower.
出处
《安阳工学院学报》
2008年第2期48-51,共4页
Journal of Anyang Institute of Technology
关键词
事务数据库
项目数据库
支持事务标识符序列
双序列串比较
Transaction database
Item database
Support transaction identifier sequence
double sequencestrings comparing
作者简介
邓铁军(1971-),男,辽宁沈阳人,辽宁林业职业技术学院信息工程系讲师。主要研究方向:网络安全及计算机通信,信息安全,多媒体信号处理。