基于有向项集图的频繁集挖掘优化算法

Updated Algorithms of Frequent Itemsets Mining Based on Directed Itemsets Graph

关联规则挖掘是数据挖掘研究中一个非常重要的研究领域。文章利用有向项集图(DISG)来存储有关频繁项集的信息，提出了利用深度优先的策略进行搜索的频繁项集挖掘的优先算法UDBDG(Updated DFS Based DISG) 。最后分析了算法在时间和空间上的复杂度并以mushroom数据库为例进行了试验。试验结果证明算法对于处理稠密集数据是有效的。

Mining association rule is a very important research field of data mining. This paper introduces a new data structure called DISG (directed itemsets graph)in which the information of frequent itemsets is stored. Based on it, a new algorithm called UDBDG(Updated DFS based DISG) is developed by using depth first searching strategy. At last, it analyses the time and space complexity of UDBDG and performs an experiment on a real dataset mushroom to test UDBDG. The experiment shows that it is efficient for mining dense datasets.