Processo de mineração de dados orientado ao usuário final: o ambiente LAWSMiner

Abstract

The complexity of extracting knowledge from the immense amount of data currently generated creates the need and brings the opportunity to develop automated mechanisms to accelerate the knowledge discovery process, without forgetting the accuracy of the results and the increase in the productivity of analysts. The knowledge discovery process consists of several sequential, well-defined and related phases, from data selection, preprocessing, mining and evaluation, until the discovery step. This work presents a proposal for automation of the data mining stage, based on a theoretical review, requirements gathering, architecture and applications study. As a contribution, an end-user-oriented data mining environment was developed and made available in the cloud. This was positively evaluated by expert users and also used in teaching-learning and self-study scenarios of Data Science.