A Precise Telecom Customer Tariff Promotion Method Based on Multi-route Radial Basis Kernel Fuzzy C-means Clustering

Precise tariff promotion for existing customers is important for telecom operators to increase their revenue. The basis of successful marketing campaign mainly relies on the analysis of users’ product usage behavior. Unfortunately, traditional segmentation methods cannot discover complex usage behavior of users, which may result in a low conversion rate during customer tariff promotion activity. To address this problem, we introduce a novel soft clustering method called Multi-Route Radial Basis Kernel Fuzzy C-Means (MRRBKFCM) for telecom customer segmentation based on their mobile internet usage data. The kernel function in the proposed algorithm was taken as the radial basis kernel function, the iterative formula for the recalculation of cluster centers was obtained by gradient method and particle swarm optimization (PSO) algorithm. Compared with the performance of the traditional Kernel Fuzzy C-means (KFCM) and Fuzzy C-means (FCM), the effectiveness of the proposed algorithm has been validated by the public UCI data sets. The customers of GuanAn branch, SiChuan telecom company are used as the input for the proposed customer segmentation algorithm. We launched the promotion campaign according to the outputs of our method for city district of Guan An, Lin Shui and Yue Chi. 31285 users has been used for customer promotion campaign. Compared with the average conversion rate \(5.81\%\) using traditional marketing strategy, our method reaches to \(19.11\%\), \(17.42\%\) and \(7.99\%\) for each district respectively, it greatly reduces marketing cost and increases the conversion rate.

[1]  Anil K. Jain.  Data clustering: 50 years beyond K-means , 2010, Pattern Recognit. Lett..

[2]  James M. Keller,et al.  A possibilistic fuzzy c-means clustering algorithm , 2005, IEEE Transactions on Fuzzy Systems.

[3]  Alex Smola,et al.  Kernel methods in machine learning , 2007, math/0701907.

[4]  P. Berger,et al.  Customer lifetime value: Marketing models and applications , 1998 .

[5]  Tao Zhang,et al.  A novel cluster algorithm for telecom customer segmentation , 2016, 2016 16th International Symposium on Communications and Information Technologies (ISCIT).

[6]  Peter C. Verhoef,et al.  Modeling CLV: A test of competing models in the insurance industry , 2007 .

[7]  S. Mohanavalli,et al.  Fuzzy Clustering for Effective Customer Relationship Management in Telecom Industry , 2011, CSE 2011.

[8]  Yan Wan,et al.  Mobile Customer Clustering Based on Call Detail Records for Marketing Campaigns , 2009, 2009 International Conference on Management and Service Science.

[9]  F. Dwyer.  Customer lifetime valuation to support marketing decision making , 1997 .

[10]  Chih-Ping Wei,et al.  Turning telecommunications call details to churn prediction: a data mining approach , 2002, Expert Syst. Appl..

[11]  Kyong Joo Oh,et al.  Facilitating cross-selling in a mobile telecom market to develop customer classification model based on hybrid data mining techniques , 2011, Expert Syst. Appl..

[12]  Mark A. Girolami,et al.  Mercer kernel-based clustering in feature space , 2002, IEEE Trans. Neural Networks.

[13]  Chui-Yu Chiu,et al.  An intelligent market segmentation system using k-means and particle swarm optimization , 2009, Expert Syst. Appl..

[14]  Stephen C. H. Leung,et al.  Segmentation of telecom customers based on customer value by decision tree model , 2012, Expert Syst. Appl..

[15]  Adem Karahoca,et al.  Comparing clustering techniques for telecom churn management , 2006 .

[16]  R S Duboff.  Marketing to maximize profitability. , 1992, The Journal of business strategy.

[17]  Joshua Zhexue Huang,et al.  Extensions to the k-Means Algorithm for Clustering Large Data Sets with Categorical Values , 1998, Data Mining and Knowledge Discovery.

[18]  Sharmila Subudhi,et al.  Use of Possibilistic Fuzzy C-means Clustering for Telecom Fraud Detection , 2017 .

[19]  Xian Fu,et al.  Kernel-based fuzzy c-means clustering algorithm based on genetic algorithm , 2016, Neurocomputing.

[20]  P. Preckel,et al.  Customer Lifetime Value: An application in the rural petroleum market , 1997 .

[21]  James C. Bezdek,et al.  On cluster validity for the fuzzy c-means model , 1995, IEEE Trans. Fuzzy Syst..

[22]  J. Bezdek,et al.  FCM: The fuzzy c-means clustering algorithm , 1984 .