Presenting a Framework for Segmentation of Life Insurance Customers Using Data Mining

The purpose of this study is to provide a framework for segmentation of life insurance customers using data mining techniques. The statistical population of the study consisted of customers of an insurance company in Isfahan, where the required data were collected from contracts of life insurance during the years 2008 to 2018. Data were collected on 353 life insurance policyholders in 14 variables (in terms of individual characteristics and selected insurance conditions). Customers were classified into 4 clusters using K-means clustering algorithm and Matlab software. The analysis of the results of the customer behavior of each cluster provided a basis for naming the clusters as progressives, conservatives, toilers, and vanguards. Also, to facilitate the process of examining the input variables in each cluster, only 7 variables (age, education, occupation, premium payment, number of beneficiaries of death insurance, death rate increase and regular premium payment) are included in each cluster. In the Chi-square test, there were two significant differences at the 0.001 level.