摘要
随着零售业在城市的快速发展,智能系统积累了大量的零售业原始数据,急需一种技术来发现数据中蕴含的内在规则,为企业管理者提供决策支持。数据挖掘是目前一个重要的研究方向,可以把日常业务数据知识化。介绍了零售业商务智能系统的发展现状,并通过分析零售业数据来掌握顾客的购买偏好,并同时对挖掘结果进行说明,在一定程度上利用关联规则技术解决现实中的商业问题。针对数量和利润的因素,提出利用频繁项目集寻找商品利润最大化的销售组合模型,零售商可以根据该模型输出的销售组合模型对商品进行捆绑销售,以获得最大利润。提出来竞争商品的概念,即找出隐含在数据库中相互竞争商品的模型,这样就得到了零售业商品推荐模型。实验结果表明,提出的模型能找出高交叉销售利润的商品,在零售业中有很好的实用性。
With the rapid development of supermarket, a lot of business data are accumulated by intelligent system. It＇ s imperative and necessary to find an effective technique to explore and discover the potential knowledge from the enormous amount of data, which is helpful for business decision making. Data mining has an important research role in the world. It can be used to acquire the knowledge. The current situation of supermarket development is analyzed,and the customer＇ s buying behavior is understood through the analysis of the retail sales data, making the explanation to the mining result, application of association rules to solve real business problems. According to the factors of the quantity and profit, the frequent item sets are adopted to find the sales combination model of profit maximization of commodity, and retailers can use it to bundling and gain the biggest profit. Based on the concept of competitive products, a model is pro- posed that can be used to find out the hidden in the retail database by the frequent and non-frequent items, getting the model of retail commodity recommendation. The experiment shows that the model can find out the high cross selling goods with good practicality in supermarkets.
出处
《计算机技术与发展》
2016年第10期146-150,共5页
Computer Technology and Development
基金
陕西省自然基金资助项目(2003JM8005)
榆林市科技局资助项目(NY13-15)
榆林学院青年科技基本资助项目(14YK37)
