数据挖掘在烟草企业CRM中的应用

发布时间：2018-04-05 20:54

本文选题：数据挖掘　切入点：客户关系管理　出处：《华南理工大学》2013年硕士论文

【摘要】：近几年，随着社会对控烟的关注度提高，控烟力度不断加大，低焦油卷烟的发展步伐逐步加快，，低焦油卷烟销售将是今后烟草销售的主要趋势。而面临着前所未有的国内外竞争压力，作为连结烟草系统与消费者桥梁的零售客户终端，可以说是决定烟草竞争力的关键。因此，如何挖掘具有发展潜力的高价值零售客户，促进卷烟零售客户经营发展向烟草行业发展方向靠拢是当前烟草网建工作的重中之重。从烟草企业的客户服务策略来看，分析型CRM是未来发展的趋势，通过对操作型CRM中的数据进行提取、分析和预测，把大量的数据转换成可靠实用的信息，指导烟草企业在卷烟销售与客户服务等方面合理配置资源，最终实现改进客户关系的目的。烟草商业企业在经营过程中积累了大量的客户订单数据，利用数据挖掘技术对客户订单数据进行分析，将客户进行细分，从而对在低焦油卷烟方面具有较大销售潜力、培育价值高的客户实施个性化服务，使客户发展与企业发展相互促进，实现客户利润与企业效益最大化。本文是基于微软SQL ServerAnalysis Service(SSAS)进行数据挖掘，主要采用基于Web的B/S体系结构，包括数据源、数据仓库、OLAP、应用服务器和客户端。根据烟草营销的实际分析需求，在SQL Server2005中建立基于零售客户卷烟销售为主题的数据仓库，且从源数据库中抽取、转换和导入相关数据到数据仓库中。接着在数据仓库上通过SSAS对分析主题建立对应的多维数据集，用DMX语言实现各种分析需求和数据的钻取、切片、切块，并用微软Reporting services开发基于web的前端数据展现。本文将研究的广州烟草某区域历史销售数据导入到SQL Server数据库中作为挖掘的数据源，并使用SSAS作为数据挖掘平台，构建数据挖掘模型，并采用决策树分类技术进行数据挖掘，实现对卷烟销售趋势的决策分析。并且，通过分析客户的销售数据挖掘卷烟零售客户的销售潜力，并对潜力客户价值进行分级排序，找出培育价值高的潜力客户。
[Abstract]:In recent years, with the increasing attention of the society to tobacco control, the intensity of tobacco control is increasing, and the development of low-tar cigarettes is gradually accelerated. The sales of low-tar cigarettes will be the main trend of tobacco sales in the future.Facing unprecedented domestic and international competition pressure, as a bridge between tobacco system and consumers, retail customer terminal can be said to be the key to determine the competitiveness of tobacco.Therefore, how to tap high-value retail customers with potential development, and promote the development of cigarette retail customers to the development direction of tobacco industry is the most important task of the current tobacco network construction.From the point of view of customer service strategy of tobacco enterprises, analytical CRM is the trend of future development. By extracting, analyzing and predicting the data in operational CRM, a large amount of data can be converted into reliable and practical information.To guide tobacco enterprises to allocate resources reasonably in cigarette sales and customer service, and to achieve the goal of improving customer relationship.Tobacco commercial enterprises have accumulated a large amount of customer order data in the course of operation. By using data mining technology to analyze customer order data and subdivide customers, tobacco commercial enterprises have great sales potential in low-tar cigarettes.Cultivate high value customers to carry out personalized service, make customer development and enterprise development promote each other, realize customer profit and enterprise benefit maximization.This paper is based on Microsoft SQL ServerAnalysis Service SSAS. It mainly adopts the B / S architecture based on Web, including data source, data warehouse, application server and client.According to the actual demand of tobacco marketing, a data warehouse based on the retail customer cigarette sales is established in SQL Server2005, and the relevant data is extracted, transformed and imported from the source database to the data warehouse.Then, the corresponding multidimensional data sets are established by SSAS on the data warehouse, and various analysis requirements and data are drilled, sliced and cut by DMX language. The front-end data display based on web is developed by Microsoft Reporting services.In this paper, the historical sales data of Guangzhou tobacco region are imported into the SQL Server database as the data source, and the data mining model is constructed by using SSAS as the data mining platform, and the decision tree classification technology is used to mine the data.To realize the decision analysis of cigarette sales trend.Furthermore, through analyzing the sales data of customers, mining the sales potential of cigarette retail customers, and ranking the potential customer value, we can find out the potential customers with high value.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.13

【参考文献】