A novel Group-FlowSort method for sorting large-scale problems with application to market segmentation

Abstract

In the big data era, promising telecom companies need to develop their user strategies when facing large-scale data. For this purpose, we considered a typical strategy decision problem like market segmentation using a novel group sorting method based on FlowSort and applied it to a real case in the telecom market. A novel Group-FlowSort procedure based on stochastic multi-criteria acceptability analysis (SMAA) was developed for sorting large-scale problems. This method could process and integrate multiple expert opinions to reach a consensus by using the Clustering In QUEst (CLIQUE) algorithm and an improved Jaya algorithm. Furthermore, Group-FlowSort based on SMAA-2 is proposed to eliminate the impact of incomplete preference information and several main characteristics and properties are discussed. In addition, a real case study including 8000 customers and 25 experts is used to illustrate the feasibility of the proposed method for telecom market segmentation. Finally, a comparative analysis with FlowSort-GDSS and sensitivity analysis using SMAA-2 are demonstrated to verify the effectiveness and robustness of the method.