基于图的挖掘关联规则改进算法

关联知识挖掘算法中一种广为人知的算法就是Aprior算法,之后所有关联规则挖掘算法的基本思想都是基于频繁项目集发现算法的基础上进行了改进.为了提高关联规则挖掘效率,首先回顾了基于图的关联规则挖掘算法;然后,在此基础上进行了改进,把关联规则挖掘中寻找频繁项集的问题转换为图中寻找完全子图的问题,通过在图中查找完全子图来寻找频繁项集.提出了一种基于图的关联规则挖掘改进算法,并且对原算法和改进的算法从时间和空间的性能进行了比较分析,得出改进的算法是有效可行的.最后从实验结果得出结论GenerateItemsets算法比DGBFIG算法优.

An Improved Algorithm for the Graph-Based Discovering Association Rules

Aprior algorithm is association rules mining algorithm which all known by the people. The outline of all Association Rules mining algorithm based on in the frequent item sets discovery algorithm in advanced. In order to enhance the association rules mining efficiency,in this paper, firstly, the graph-based discovering algorithm with association rules is discussed. Then, the author modifies this algorithm, transform the problem of find frequent item sets to discovering the clique graphs in the graph ,by searching the clique graphs in the association graph to find frequent itemsets. In this paper,author proposes an improved algorithm for the graph-based discovering association rules,and we compare the improved algorithm with the former algorithm from the complexity of times and spatial. Conclude in this algorithm efficiency. Finally, we see the GenerateItemsets algorithm prior to DGBFIG algorithm from experimental results.