【作者】 张震宇；

【导师】 杨丹；

【作者基本信息】 重庆大学 ， 软件工程， 2004， 硕士

【摘要】 计算机技术发展的历史在某种意义上也是数据自动采集、生成和处理的历史。目前，全球为各行各业服务的数据库不仅在数量上快速增长，而且在规模上也越来越大。数据挖掘正是在这样的应用需求背景下产生并迅速发展起来的一个重要的研究领域。随着垄断格局的打破，保险公司如何应对激烈的市场竞争？如何培育公司的核心竞争力？方法只有一个：以市场为导向，以客户为中心，细分客户，建立完善的客户服务管理体系，实施品牌制胜战略。客户关系管理系统（即CRM）是现代经营管理科学与现代信息技术结合的产物。本文主要着重于数据挖掘技术在保险业CRM系统中的应用研究。主要采用了分类树和关联规则方法，来发现隐含在客户行为中的规则，将抽象的数据转换成用户可以利用的信息。并对关联规则中的增量更新问题进行了分析，提出了在最小支持度和最小可信度都不变的情况下解决方法。并将CRM中关于客户关系的理念应用到数据仓库中，提出了如交叉销售、风险客户、客户流失等数据模型。从而细分客户、细分服务，实现经营和服务的差异化，以此提高保险公司的决策支持和商业智能水平。更多还原

【Abstract】 To some extent, computer technology evolvement includes the evolvement of data To some extent, computer technology evolvement includes the evolvement of data extraction, data generation and data processing technology. Currently, operational database in all industry fields grows rapidly not only in number, but also in data volume. Data mining is a major research area, which is driven by the application requirement in this field. After monopoly in insurance service industry is broken, strategies for insurance c o mpany to get survived in the competitive market, include market- driven/ customer- oriented concepts, customer classification in finer granularity, comprehensive customer service management system, and building strong service brand. Customer relationship management (CRM) system is the application of information technology in modern business management.Decision tree and rule-based method are adopted to explore the possible rules in customer business behavior and transform the raw data into useful information. Additionally, the incremental update issues have been discussed and a solution is presented without the change of minimum supporting degree and minimum trust degree. Furthermore, the concept of customer relationship in CRM is applied into data warehouse, and build the models of overlapped sales, risking customer and customer losing. Therefore, the customers and services are defined in more details and implement the difference of operation and services in order to improve the decision making and business intelligence.更多还原