Customer base segmentation of a telecommunication company by using Bayesian semi-parametric analysis of traffic data

Customer segmentation is one of the most important purposes of the customer base analysis for telecommunication companies. Since companies accumulate very large amount of data on costumer behavior, segmentation is typically achieved by profiling and clustering traffic behavior jointly with demographic data and contracts characteristics. Unfortunately most algorithms and models applied for segmentation does not take into account the longitudinal characteristics of data. In particular in telecommunication traffic analysis, it is well known the importance of decreasing pattern of traffic in customer life and it is relevant to aggregate all clients with such a pattern, while other unknown clusters may be of interest of the marketing manager. Our approach for analyzing such a problem is based on specifying the distribution of functions as a mixture of a parametric hierarchical model describing the decreasing pattern segment and a nonparametric contamination that allows unanticipated curve shapes in subjects traffic. The parametric component is chosen based on prior knowledge, while the contamination is characterized as a functional Dirichlet process.