Modelo de caracterización del perfil del cliente adaptado al sector de la construcción, aplicando minería de datos: caso de estudio en empresa constructora

Abstract

Building houses process in construction companies is affected by fallen negotiations before delivery of the living place, unknowing the reasons for this situation. In building sector there is no a model that allows the characterization of customer profile linked to the commercialization process that would contribute to mitigate the risk to business agreement. As several productive sectors begin to apply data mining, adaptations of traditional methodologies for data mining arise, which solve particularities of these sectors. This paper proposes a characterization model of customer profile adapted to building sector with data mining that includes the results obtained in case study in a building company. The Knowledge Discovery in Databases (KDD) methodology was followed with an adaptation or contribution found in literature that identified own characteristics of building industry in the approach, design, mining techniques, pattern evaluation and validation that allowed generalizing the model. With case study, a design for the data warehouse based on the multidimensional data model using the star schema was proposed. It was proved that classification, regression and clustering techniques are recommended to solve the problem and finally the proposal for data mining techniques evaluation and validation was presented. The generalization of the characterization model of customer profile, confirmed the tendency and requirement of the productive sectors to have data mining models specific to their needs.