| 摘要: |
| 文章分析了在KDD研究中现有的关联规则发现算法关于频繁项目集的生成与测试方法,提出了一种新的基于垂直数据分布的关联规则发现算法.该算法无需复杂的Hash数据结构,仅需对整个数据库作两次遍历,从而既方便了实现,又提高了效率. |
| 关键词: 关联规则,频繁项目集,等价类. |
| DOI： |
| 分类号: |
| 基金项目:本文研究得到国家自然科学基金和国家教育部博士点基金资助. |
|
| An Efficient Algorithm for Discovering Association Rules Based on Vertical Data Layout |
|
OU-YANG Wei-min,CAI Qing-sheng
|
| Abstract: |
| In this paper, the authors analyze the methods to generate and test frequent itemsets in existing algorithms in KDD research, and put forward a new efficient algorithm for discovering association rules based on vertical data layout. This algorithm has no need of Hash data structure and makes only two databases scans. As a result, the algorithm not only facilitates the implementation, but also improves the efficiency. |
| Key words: Association rule, frequent itemsets, equivalence class. |
