Identification of overpricing in the purchase of medication by the Federal Government of Brazil, using text mining and clustering based on ontology

Abstract:

Increasing the transparency level in his actions and spending is one of the primary duties of the Brazilian Federal Government. The creation of laws that oblige full disclose of all its expenditures through transparency portals enables citizens to supervise all government entities. However, only the dissemination of these data, without a definite standard or the availability of data analysis tools, does not guarantee that the citizen is empowered to play his role. Therefore hence, the objective of this work is to identify overprice in the acquisition of products purchased by the federal government of Brazil using the unstructured data available on the Transparency Portal. The last two-years’ worth of purchasing data, available in the Transparency Portal, were extracted, processed and stored. Due to his diverse nature and high volume of data, this study focused only on medicines purchased by the Ministry of Health. Ontology-based text mining and clustering techniques were applied for automatic identification and classification of products. The processing of this information was done through text mining and clustering, based on the ontology registered in another database of the Brazilian government. Because of this work, a consolidated price base per medication was created to allow the identification of distortions in prices practised, facilitating the identification of cases that merit further investigation to unravel fraud to the treasury.


Reference:
CORREA, Marco Aurelio Oliveira dos Santos; LEAL, Adriano Galindo. Identification of overpricing in the purchase of medication by the Federal Government of Brazil, using text mining and clustering based on ontology. In: INTERNATIONAL CONFERENCE ON CLOUD AND BIG DATA COMPUTING, ICCBD, 2., 2018, Barcelona. Proceedings… 5 p.

Document is password protected, ask Customer Service/Library-DAIT/IPT. Log into BiblioInfo Biblioteca-DAIT/IPT to access the text in PDF:
https://escriba.ipt.br/pdf_restrito/175621.pdf

SUBSCRIBE to our newsletter

Receive our news in your email.

INSCREVA-se em nossa newsletter

Receba nossas novidades em seu e-mail.