基于GSP算法的卷烟消费者研究  

Consumer Research Based on Data Mining Algorithm

作　　者：黄飞杰[1] 张卫东[2] 侯石鹏 宋红文[2] HUANG Feijie;ZHANG Weidong;HOU Shipeng;SONG Hongwen(Internet Research Center,Guangxi Zhongyan Industry Co.,Ltd.,Nanning Guangxi 530000,China;Southwest University of Science and Technology,Mianyang Sichuan 621010,China)

机构地区：[1]广西中烟工业有限责任公司互联网研究中心,广西南宁530000 [2]西南科技大学,四川绵阳621010

出　　处：《信息与电脑》2022年第16期58-60,共3页Information & Computer

摘　　要：烟草专卖法规定,卷烟消费购买只能在线下进行,导致收集卷烟消费者数据困难。当卷烟行业进行品牌营销活动时,无法精准找到目标消费人群,营销效果不佳。因此,基于广义序列模式(Generalized Sequential Pattern,GSP)算法,先采用滑动时间窗技术构建消费者序列表,再构建频繁集,最后通过GSP算法挖掘卷烟消费者的购买行为规律。通过挖掘真龙(美人香草)消费者行为揭示了这类消费者的购买来源和去向品牌,基于此规律开展品牌营销活动,实现精准营销的活动效果。The Tobacco Monopoly Law stipulates that cigarette consumption can only be purchased offline,which makes it difficult to collect cigarette consumer data.When the cigarette industry conducts brand marketing activities,it is unable to accurately find the target consumers,and the marketing effect is poor.Therefore,based on the Generalized Sequential Pattern(GSP)algorithm,we first use the sliding time window technology to build the consumer sequence table,then build the frequent set,and finally use the GSP algorithm to mine the rules of cigarette consumers'purchase behavior.By digging into the consumer behavior of Zhenlong(Mermaid's Vanilla),the purchase source and destination brand of such consumers are revealed.Based on this rule,brand marketing activities are carried out to achieve the effect of precision marketing activities.

关 键 词：广义序列模式(GSP)算法 滑动时间窗 消费者 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]