论文部分内容阅读
本文调研收集了顾客在全品类家电卖场购物过程中逗留产品区域、所购产品、顾客特征信息。对每条顾客记录中购买产品向量(设所有产品构成一个高维空间,顾客购买某产品就记为1,未买某产品记为0,产品购买记录是一个高维向量)和逗留产品区域向量按变量进行聚类,得到以产品组合为分类标准的业态结构。均值检验不同业态下的顾客特征,若顾客异质,说明这种业态结构从顾客细分角度来说是合理的。
This article collects the customer’s product information, product information and customer’s characteristics during the shopping process of the whole home appliance store. For each customer record, a product vector is purchased (assuming that all the products form a high-dimensional space, the customer purchases a certain product as 1, the non-purchased product as 0, and the product purchase record as a high-dimensional vector) According to the variables, clustering is carried out to obtain the format structure based on the product combination. The mean tests customer characteristics under different formats. If the customers are heterogeneous, it shows that this format is reasonable from the perspective of customer segmentation.