With the development of technology, people have entered the Big Data era, where the insurance companies meet with opportunities for their massive customer data. To make full use of these data and promote the insurance companies’ benefits, this paper aims to propose a method to reasonably measure the similarity among customers and segment them. However, there exists a challenge to deal with different types of data, i.e., linguistic data and crisp numbers, in customer relationship management. Considering that probabilistic linguistic term set possesses an apparent superiority in reducing linguistic information losses during the process of customer evaluation and data processing, we give a novel correlation measure of two data sequences with both crisp numbers and probabilistic linguistic term sets. Furthermore, we put forward the weighted form of the correlation measure to increase its application elasticity, and then apply the proposed correlation measures to insurance companies’ customer relationship management based on an improved clustering algorithm. Finally, a numerical example is given to illustrate the validity of our correlation measures.
