Application of data mining techniques in the characterization of internal personnel turnover

The aim of this work is to propose the application of data mining techniques to detect behavior patterns in employee management databases. The desired knowledge concerns the transfer of people between organizational areas (internal turnover), in order to provide organizations a model for assessing the impact of these transfers on their people management policies. Therefore, this article describes a model that applies techniques of clustering and association rules for the discovery of knowledge on transfers of staff between units of a Brazilian financial organization. To allow the expression of this knowledge in a compact expressive way, a new dimension type is proposed for the multidimensional model. This way, the hypothesis is evaluated on a case study conducted in a Brazilian organization, with over four thousand units spread nationwide and over 95 thousand employees.