Bibliographic citations
Baldoceda, J., Prati, J. (2017). Modelo de clasificación de clientes con telefonía móvil y uso de canal de autogestión USSD en una empresa de telecomunicaciones [Tesis, Universidad Nacional de Ingeniería]. http://hdl.handle.net/20.500.14076/10934
Baldoceda, J., Prati, J. Modelo de clasificación de clientes con telefonía móvil y uso de canal de autogestión USSD en una empresa de telecomunicaciones [Tesis]. : Universidad Nacional de Ingeniería; 2017. http://hdl.handle.net/20.500.14076/10934
@misc{renati/708525,
title = "Modelo de clasificación de clientes con telefonía móvil y uso de canal de autogestión USSD en una empresa de telecomunicaciones",
author = "Prati Oblitas, Jean Pierre",
publisher = "Universidad Nacional de Ingeniería",
year = "2017"
}
This thesis develops a classification model using bayesian network predictive techniques to identify clients that do not use the USSD channel, a channel currently used by telecommunications companies for their customer service. The proposed solution to classify clients covers the development of a datamart, the application of bayesian data mining algorithms and the model applying the algorithm selected. The first chapter describes the problem in the customer service area of telecommunications companies, the need to classify their customers by their behavior using the care channels that the company offers them; indicates the objectives, justification and scope of the thesis. The second chapter is the theoretical framework and presents the concepts necessary to understand the development of the thesis. This includes three important points for the solution we propose: datamart, data mining and care channels. The third chapter shows the general outline of the proposed solution and the selection of methodologies and tools to be used in the three phases of the solution. The fourth chapter covers the development phase of the datamart, uses the methodology proposed by Ralph Kimball as a reference and defines the information needs for the solution. In fifth chapter we develop the classification model following the CRISP-DM methodology, we use the datamart that we presented in the previous chapter as data source. The model is constructed for three classification algorithms of Bayesian networks. The last two chapters comprise the tests of the models with the three classification algorithms, we present the results of each one test for the model selection. The final part is the conclusions reached after the tests of the classification model and recommendations for similar projects. Keywords: classification model, USSD self-management channel, datamart, data mining, bayesian networks, predictive model, CRISP-DM, Kimball methodology.
This item is licensed under a Creative Commons License