加权频繁项集挖掘算法在超市商品捆绑销售中的应用

The Application of Weighted Frequent Itemsets Mining Algorithm in Binding Sales of Supermarket

引入一种新的加权关联规则支持度和置信度的计算方法,并利用矩阵的存储结构提出一种新的加权关联规则挖掘算法,从而改进了加权频繁项集的挖掘效率.该算法在Apriori算法的基础上,对数据库仅需扫描一次,能很快地计算项集的支持度,大大减少了I/O次数,有效提高了加权频繁项集的生成效率.通过应用于超市捆绑销售,证明了该算法能有效地提取商品间的关联信息,有助于商品的销售.

Introducing a new calculation of weighted support and confidence,this paper proposes a weighted association rule mining algorithm based on matrix,which can improve the weighted frequent itemsets mining methods.Based on Apriori algorithm,it used matrix＇s storage structure,which scans the database only once and can quickly calculate the supporting degree of itemsets.Therefore,this algorithm greatly reduces the I / O,and can improve the weighted frequent itemsets generation effectively.By being used in the binding sales of supermarkets,shows that the algorithm can extract the relationship between product information and help the sales of goods effectively.