Rastro-DM: data mining with a trail
It addresses documentation gaps in data mining for institutional use, though it is incremental as it complements existing methodologies like CRISP-DM.
The paper tackles the problem of documenting data mining projects by proposing Rastro-DM, a methodology focused on tracking processes rather than models, and illustrates its application in a project for classifying PDF documents in Brazilian public treasury investigations.
This paper proposes a methodology for documenting data mining (DM) projects, Rastro-DM (Trail Data Mining), with a focus not on the model that is generated, but on the processes behind its construction, in order to leave a trail (Rastro in Portuguese) of planned actions, training completed, results obtained, and lessons learned. The proposed practices are complementary to structuring methodologies of DM, such as CRISP-DM, which establish a methodological and paradigmatic framework for the DM process. The application of best practices and their benefits is illustrated in a project called 'Cladop' that was created for the classification of PDF documents associated with the investigative process of damages to the Brazilian Federal Public Treasury. Building the Rastro-DM kit in the context of a project is a small step that can lead to an institutional leap to be achieved by sharing and using the trail across the enterprise.