Characterizing Educational Data Mining

Educational data mining (EDM) is intended to uncover trends and patterns hidden in the thousands of data sets in an educational system and has drawn the attention of academic authorities for being able to bring benefits to educational institutions. This work aims to investigate in more detail the mining of educational data and to suggest a preliminary classification scheme. As a result, we hope to provide an overview of such research area by identifying key topics, types and trends of preliminary research, as well as the maturity of existing contributions. Since the EDM is part of an interdisciplinary area, mobilizing mainly knowledge of statistics, machine learning, pattern recognition, etc., we believe that with the present work one can have a better understanding of the area and its aspects.