Classificação e clustering aplicados à licitações

Abstract

The bidding processes are very important to ensure that public resources are well spent. In Goiás, at the state level, the transparency portal offers a series of options for data transparency, which allows the analysis of this data, as well as social control tools for accountability in a very transparent way. In this work, data mining is used, both in the form of classification and grouping, to obtain hidden trends in large volumes of data in the bidding area. A study is carried out using the Weka software (Waikato Environment for Knowledge Analysis) that uses decision tree techniques and neural networks analyzing the bidding modalities to find trends that can improve the decisions of public managers.