动态数据库关联规则增量更新算法研究

The Updated Algorithm of Increment for Association Rules Based on Dynamic DataBase

模式维护是数据挖掘中一项具有挑战性的任务。针对大部分关联规则,更新算法只考虑最小支持度这一因素,没有考虑最小置信度阀值;数据库更新时只考虑数据添加,没有考虑数据的删除。提出一种可同时考虑以上各方面的动态数据库更新算法,可有效挖掘出感兴趣的知识。实验表明,该算法切实可行。

Pattern maintenance is a challenging task in data mining. As the majority updated algorithms for association rules have the same shortcomings which only take account of the minimum support and increase data, without the minimum confidence and decrease data. In this paper, a new updated algorithm for dynamic database which could avoid some shortcomings of the above. It can effectively mine some interested knowledge. The experiment shows that the proposed approach is efficient.