Data science to identify crimes against public administration

Abstract

Context:The management of public resources is subject to illegal acts and the automatic identification of such acts depends on the analysis of a lot of data. Objective: The object of this work is the analysis of scientific publications through a study based on systematic mapping with the purpose of evaluating them in relation to the use of automated tools to identify crimes against public administration in databases from the perspective of researchers in the data science context. Method: Using PICO strategy (Population, In-tervention, Comparison, and Outcome), a systematic mapping was conducted to find the primary studies in the literature and collect evidence for directing future research. Results: Nineteen works were found that fit the proposed cri-teria. Almost 80% of the studies found seek to identify some type of fraud in bidding processes, obtaining accuracies between 72% and 99%. The research also revealed different techniques for approaching the problem. Considering all the works, the most used databases are bidding bases, lawsuits, public notices and corporate structure of companies, respectively. Conclusions: The work has shown a recent increase in interest in analyzing public data for irregularities. It is expected that this analysis will help control bodies elucidating different ways of detecting crimes against the public administration in an automated way.