| 一种基于频繁序列树的增量式序列模式挖掘算法 |
| |
| 引用本文： | 刘佳新.一种基于频繁序列树的增量式序列模式挖掘算法[J].计算机与现代化,2012(2):8-10,18. |
| |
| 作者姓名： | 刘佳新 |
| |
| 作者单位： | 燕山大学图书馆,河北秦皇岛066004 |
| |
| 基金项目： | 国家自然科学基金资助项目（61170190）; 秦皇岛市科学技术研究与发展计划项目（201001A018） |
| |
| 摘 要： | 针对目前现有的增量式序列模式挖掘算法没有充分利用先前的挖掘结果,当数据库更新时,需要对数据库进行重复挖掘的问题。本文提出一种基于频繁序列树的增量式序列模式挖掘算法(ISFST),ISFST采用频繁序列树作为序列存储结构,当数据库发生变化时,ISFST算法分两种情况对频繁序列树进行更新操作,通过遍历频繁序列树得到满足最小支持度的所有序列模式。实验结果表明,ISFST算法在时间性能上优于PrefixSpan算法和IncSpan算法。
|
| 关 键 词： | 数据挖掘 序列模式 增量式挖掘 频繁序列树 投影数据库 剪枝策略 |
An Incremental Mining Algorithm of Sequential Patterns Based on Frequent Sequence Tree |
| |
| LIU Jia-xin.An Incremental Mining Algorithm of Sequential Patterns Based on Frequent Sequence Tree[J].Computer and Modernization,2012(2):8-10,18. |
| |
| Authors: | LIU Jia-xin |
| |
| Affiliation: | LIU Jia-xin(Library of Yanshan University,Qinhuangdao 066004,China) |
| |
| Abstract: | This paper proposes an incremental mining algorithm of sequential patterns based on frequent sequence tree,called ISFST,in order to solve the problem that the existed incremental mining algorithms can not make full use of the results of the previous mining,when the database is updated,the algorithms need to mine the database once again.ISFST uses the frequent sequence tree as the storage structure of the algorithm.When the database is updated,ISFST is divided into two kinds of situations to update the frequent sequence tree,and finally gets all sequential patterns.Experiments show that ISFST outperforms PrefixSpan and IncSpan in time cost. |
| |
| Keywords: | data mining sequential patterns incremental mining frequent sequence tree projected database pruning strategy |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
