| 基于PrefixSpan 序列模式挖掘的一种改进算法 |
| |
| 引用本文： | 吴楠,胡学钢.基于PrefixSpan 序列模式挖掘的一种改进算法[J].数字社区&智能家居,2007(20). |
| |
| 作者姓名： | 吴楠 胡学钢 |
| |
| 作者单位： | 合肥工业大学计算机与信息学院,安徽,合肥,230009;宿州学院计算机科学与技术系,安徽,宿州,234000 合肥工业大学计算机与信息学院,安徽,合肥,230009 |
| |
| 摘 要： | PretixSpan算法解决了类Apriori算法的不足,但产生的投影数据库花费了较多的存储空间及扫描时间.本文基于PretixSpan算法提出PSD算法,舍弃了对非频繁项的存储及对投影序列数小于最小支持数的投影数据库的扫描,减少了不必要的存储空间,提高了查询速度.实验证明,PSD算法比PretixSpan算法具有更好的时空性能.
|
| 关 键 词： | 数据挖掘 序列模式 投影数据库 |
An Improved Algorithm for Mining Sequential Patterns based on PrefixSpan |
| |
| WU Nan,HU Xue-gang.An Improved Algorithm for Mining Sequential Patterns based on PrefixSpan[J].Digital Community & Smart Home,2007(20). |
| |
| Authors: | WU Nan HU Xue-gang |
| |
| Affiliation: | WU Nan1,2,HU Xue-gang1 |
| |
| Abstract: | PrefixSpan solved the shortages of Apriori-like algorithms. But projection databases cost more redundant memory and scan-time. A sequential pattern mining algorithm(PSD) based on PrefixSpan is presented in this paper. The algorithm decreases the redundant memory and scan-time by abnegating the non-frequent items and projection databases which sequential number is lower than minimum support. Experiments demonstrate that PSD is more efficient than PrefixSpan. |
| |
| Keywords: | data mining sequential pattern projection database |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
