关联规则及其在商品销售中的应用研究

发布时间：2018-05-28 11:48

本文选题：关联规则 + 商品销售　；参考：《湖北大学》2014年硕士论文

【摘要】：随着数据挖掘技术不断发展,数据挖掘带来的商业价值越来越受到各行业的关注,急切想借助数据挖掘东风创造更多价值。衡量一个企业创造更多价值的直接标准是其商品销售额和销售量的增加。关联规则是数据挖掘应用最广泛的算法之一,关联规则在商品销售中的应用虽然很早就有成功案例,但随着商业竞争的日益激烈,如何提高商品推荐的有效性成了亟待解决的难题。围绕提高商品销售的有效性问题,本文主要从以下几个方面开展工作：首先,在广泛查阅有关资料的基础上,深入研究了已有的关联规则挖掘技术和聚类方法。在此基础上,通过大量实验,选取出适合商品销售推荐的关联规则挖掘算法和聚类算法；其次,利用选出的聚类算法,提取用户特征,对客户进行分类；在分类客户数据的基础上,将商品销售数据集进行转换,形成矢量化的交易数据库,便于后期挖掘效率的提高；然后,利用选出的关联规则挖掘算法对转换数据库进行挖掘,得出针对不同类型客户的挖掘结果。基于上述思路,本文提出了一个CAM(Cluster-Association Mining)算法；最后,将提出的算法在真实数据集上予以了实现。结果表明,本文提出的算具有较强的针对性和有效性。
[Abstract]:With the development of data mining technology, the commercial value brought by data mining is paid more and more attention by various industries, and it is eager to create more value with the help of data mining. The direct measure of an enterprise's creation of more value is the increase in its merchandise sales and sales. Association rules is one of the most widely used algorithms in data mining. Although the application of association rules in commodity sales has been successful for a long time, but with the increasingly fierce business competition, How to improve the effectiveness of commodity recommendation has become a difficult problem to be solved. Focusing on improving the effectiveness of commodity sales, this paper mainly from the following aspects of work: Firstly, the existing association rule mining techniques and clustering methods are studied on the basis of extensive reference to relevant data. On this basis, through a large number of experiments, the association rules mining algorithm and clustering algorithm suitable for commodity sales recommendation are selected. Secondly, the selected clustering algorithm is used to extract the user features and classify the customers. On the basis of the classified customer data, the commodity sales data set is transformed to form a vectorized transaction database, which is convenient to improve the efficiency of mining in the later stage. Then, mining the conversion database with the selected association rules mining algorithm, and obtaining the mining results for different types of customers. Based on the above ideas, a CAM(Cluster-Association mining algorithm is proposed in this paper. Finally, the proposed algorithm is implemented on the real data set. The results show that the proposed algorithm has strong pertinence and effectiveness.
【学位授予单位】：湖北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F721;F224

【参考文献】