基于改进LRFMC模型的航空公司客户分类

Airline Customer Classification Based on Improved LRFMC Model

摘要: 随着人们生活水平的提高，越来越多的人会选择飞机交通工具出行。对于同一个航线，人们可以从多个航空公司的航班之间进行选择，这一行为使得各个航空公司的竞争不断加剧。如何保留住老客户的同时，还增加新客户的数量成为了各个航空公司需要解决的问题，客户关系管理是其基础任务。客户关系管理首先需要将客户根据不同特性划分成多个类别。本研究将基于国内某航空公司客户的数据对LRFMC模型进行调整和改进。为了达到降低LRFMC模型中指标的共线性和完善的目的，将原模型中的飞行总公里数这一指标替换成平均每次飞行公里数并增加每公里机票票价构建新模型。使用层次分析法计算新模型中各指标的权重，最后通过K-Means算法对客户数据进行分类。根据分类的结果，航空公司通过对不同价值的客户采取不同的方式和策略来提高其公司的收益。

Abstract: With the improvement of the level of life, more and more people will travel by airplane. For the same route, people can choose it from multiple airlines’ flights, which makes the competition among airlines. How to keep customers and increase the number of customers has been a problem that airlines need to solve, and customer relation management (CRM) is the basic task. CRM needs to classify customers into multiple categories based on their characteristics. This study will adjust and improve LRFMC model based on the data of airline customers. In order to reduce the covariance in LRFMC model, the new model is constructed by replacing the index of total kilometers with average kilometers per flight and adding average fare per kilometer. The weights of each indicator in the new model are calculated by using Analytic Hierarchy Process (AHP), and finally the customers’ data are classified by K-Means algorithm. Based on the results of classification, airlines can adopt different approaches and strategies for customers with different values to improve revenue.