Title: Mineração de Dados Aplicada a CRM em uma Base de Clientes de Telefonia de Longa Distância
Authors: Azevedo, Hugo L. C.; Vellasco, Marley B. R.; Passos, Emmanuel P. L.
Abstract: Nowadays, it is very important for retail companies to understand their customers and create relationships with them. It is then crucial to be able to segment the customers according to their buying patterns and needs. Unfortunately, this is not an easy task and several mathematical algorithms and other techniques are required to accomplish it. In this work, a database of a long distance carrier company containing client-calling patterns was clustered and characterized. First, the Kohonen algorithm was used to cluster a subset of the database. The clusters found were then characterized using rules and some multivariate statistical visualization techniques. Finally, a classification model was built to classify future costumers and the ones who were out of the dataset originally clustered.
Paper as PDF: 5cbrn_085.pdf
BibTex file: 5cbrn_085.bib