基于划分和压缩数据库的改进Apriori算法

针对Apriori算法在面对大规模数据时效率较低的问题,提出了一种基于划分和压缩数据库的改进方法。该方法首先依据特征数据出现的频率将数据按照升序存储在临时数组中;然后将原始事务数据库分为几个互不相交的事务数据库,使得子数据库能够容纳在内存中;最后根据每个子数据库计算出的频繁项集计算整个数据库的频繁项集,从而消除了不必要的冗余数据。通过改进可以将大规模数据集进行有效的划分和压缩,对子数据库进行关联规则挖掘。实验结果表明,改进的Apriori算法在针对海量数据挖掘的执行速度和效率都有很大提高。

Improved Apriori algorithm based on classification and database compression

When the Apriori algorithm faces massive data,its rate is low.To counter the above problem,this paper puts forward an improved method based on the classification and database compression.Firstly,according to the appearing frequency of characteristic data,this method stores the data in a temporary array in ascending order.Then the original transaction database is divided into several disjoint transaction database in order to accommodate the daughter database in the memory.At last,the entire database frequent itemsets are calculated by the frequent itemsets calculated according to each daughter database, thereby eliminating the unnecessary redundant data. Through the improvement,the large data sets can be effectively divided and compressed,and the association rules can be tapped on the daughter database.The experimental results show that the improved Apriori algorithm has improved a lot in the speed and efficiency of mining the massive data.