摘要
为了从现有的CRM系统中发现潜在的车辆保险客户,提出了一种基于粗糙集理论的数据挖掘方法。利用粗糙集在知识系统中特有的分类特性,经对数据的预处理,较好地解决了样本数据中存在的属性不均衡及反向样本给数据挖掘带来的负面影响。以著名的The Insurance CompanyBenchmark(COIL 2000)作为测试数据集,经编程构建了数据挖掘模型,对客户社会背景和保单数据进行了综合挖掘测试。结果表明利用粗糙集理论对知识的分类能力,可以挖掘出数据集中潜在的对车辆保险感兴趣的客户,并给出样本分类的包含度。
In order to identify potential policyholders of vehicle insurance in customer relationship management（CRM） system,this paper presents a method for mining the data based on rough set（RS） theory.After the original dataset was cleaned by using classification characteristics of RS in the knowledge system,the problem was solved that negative impact on data mining caused by the unbalanced and contradictory of samples.Taking The Insurance Company Benchmark（COIL 2000） as testing data,data mining model was established,and the comprehensive data mining test of customers′ background and policy data was done.The results show that using the knowledge classification of RS theory,we can find out the potential customers in the vehicle insurance from data set,and give containing degrees for the classification of samples.
出处
《沈阳航空航天大学学报》
2013年第3期65-67,共3页
Journal of Shenyang Aerospace University
关键词
车辆险
数据挖掘
粗糙集
保险CRM系统
vehicle insurance
data mining
rough set
insurance CRM system
