Mineração de dados aplicada em processos fiscais

Abstract

This work intends to apply data mining techniques within the scope of fiscal representations, with the objective of predicting and classifying crimes against public order and the duration of administrative proceedings, administrative to the contribution of the public power in obtaining a decision. A bibliographical review was carried out with knowledge of knowing taxation, the processes of tax representations for criminal purposes, data mining and mining techniques. A survey of data from public representations available by the Federal Revenue was carried out, in which the year 2018 to 2020 was extracted. Then, the selection, cleaning and organization of data manually, with the aid of an electronic spreadsheet. Afterwards, the WEKA software was used to perform the data mining, with the intention of its output and distribution results, to predict and classify the data that have the potential to contribute to the public authorities in the improvement and optimization of processes.