基于关联规则的超市购物个性化推荐研究

Research on Personalized Recommendation of Supermarket Shopping Based on Association Rules

【目的】优化因消费者需求改变而受到局限的传统方法。【方法】通过收集沃尔玛超市两个时段9835名顾客的消费数据,使用Apriori算法对数据进行挖掘和分析,得到3种推荐方法,分别为按商品销售排名推荐最畅销的前N件商品、根据Apriori算法挖掘出的关联规则进行商品捆绑销售、根据定义的强关联规则推荐系数来推荐商品。【结果】在运行程序后,发现后两种推荐方式挖掘出的关联规则商品不完全相同,且对相同关联规则商品的推荐度不同。【结论】推荐策略的实施将提高超市商品的销售量,有助于满足顾客的个性化需求。

[Purposes]The aim of this paper is to optimize traditional methods that are limited by changes in consumer demand.[Methods]By collecting the consumption data of 9835 customers in Wal-Mart supermarket in two periods,Apriori algorithm was used to mine and analyze the data,and three recommendation methods were obtained,which were recommending the best-selling top N items according to the sales ranking of goods,bundling goods according to the association rules mined by Apriori algorithm,and recommending goods according to the defined strong association rule recommendation coefficient.[Findings]After running the program,it is found that the association rule products mined by the latter two recommendation methods are not exactly the same,and the recommendation degree of the same association rule products is different.[Conclusions]The implementation of the recommendation strategy will increase the sales of supermarket goods and help to meet the personalized needs of customers.