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 Company Benchmark (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。

Recently,with the rapid development of civil aviation,the scale of airlines continues to expand,and all airlines have carried out a comprehensive dispatchcentred operation control.The role of flight dispatch has become increasingly important.Today,with dramatic increases in customers and trade,aircrafts are also able to adapt to the special conditions at high altitude airportsplateau.On the other hand,due to passenger demand and airline development,missions have to be carried out in topographical conditions in the highlands.The poor flight conditions at high altitude airports of ask for higher requirements from flight dispatchers.This paper discusses the characteristics of plateau airports and routes,and analyzes the special aircraft performance at the high altitude airports and the effect on the dispatch release under conditions of the plateau airports.Finally,characteristics of the dispatch release are summed up with the examples of the Shuangliu Airport in Chengdu and Huanglong Airport at Jiuzhaigou.