一种数据挖掘中的冰山查询优化算法

An Optimization Algorithm of Iceberg Query in Data Mining

在数据挖掘技术中,关联规则可以挖掘发现大量的数据中项集之间有趣的关系或相关联系。冰山查询是关联规则挖掘的一种算法,特别是针对购物篮分析。介绍了数据挖掘的具体应用,使用改进的BUC算法(BUCE算法)以深度优先的处理方法,并综合其他传统算法,提高了处理冰山查询的有效性。同时BUCE算法在数据分组后聚集,减少I／O开销;对分组的数据排序,有利于数据的剪枝处理,将小于最小支持度的分组以后的数据可以全部剪去。

Association rules can mine a large mount of interesting relations and associations among data item sets. Iceberg query is a mining algorithm based on association rules, especially for market rules analysis. Some particular applications of data mining and focuses were introduced on the enhancing BUC optimization algorithm (BUCE). Combining a depth priority processing method with other traditional algorithms, BUCE improves the efficiency of Iceberg query. Clustering after grouping data, the BUCE algorithm decreases the processing cost. For data pruning processing, the grouped data are sorted. The data with support degrees less than the minimal support degrees can be deleted. The BUCE algorithm raises efficiency of execution of BUC.