From the textual description of the dataset available at one of the links provided in the submission, I believe that the dataset does, in fact, contain addresses:
O DNE é um banco de dados que contém mais de 900 mil CEP de todo o Brasil, constituído de elementos de endereçamento (descrição de logradouros, bairros, municípios, vilas, povoados) e Códigos de Endereçamento Postal - CEP.
However, item B of question B3 (“Addresses”) was left unchecked. Maybe that information has been overlooked.
I found another dataset that might qualify better for Locations in Brazil. It’s the “Cadastro Nacional de Endereços para Fins Estatísticos” (National Addresses Registry for Statistical Purposes), released by IBGE. It is open and free of charge, and has all the addresses in Brazil, which IBGE does use in order to conduct the national census. It has neither ZIP codes nor geolocation, but otherwise I think it would qualify better than the Correios one.
Links to downloads of TXT files (fixed width delimited tables), separated by state
A map with a visualization of statistics of this database
License, considered to be open (public domain), the same as the other datasets offered by IBGE
Ease of use is not great, as the fixed width tables require more labour to work with than, say, CSV. Also, it is not available in bulk, at least not that I could find. But being free, in contrast with the Correios one, is a huge advantage.
Considering that, we might want to replace the dataset considered in the GODI survey.
I am using the Universal Postal Union to determine what is the official gov organization in charge with postal codes. According to the UPU, the responsible gov entity is the Ministério das Comunicações and the designated operator is CORREIOS (http://www.upu.int/en/the-upu/member-countries/americas/brazil.html). Correios does not make the data available online, but only an online search service. The data from IGBE could be very useful, but it is not compliant with GODI requirements. However, I think it is important to put this link in the comments.
In that case, I think you are suggesting criteria to compare datasets and evaluate which one qualifies better for the GODI, right? So, if I have an incomplete open not-that-easy-to-use dataset would be better than a complete (supposedly easier to use) that is restricted, right?
I do agree with that criterion, but we thought that Correios had the only available dataset for that information.
Here’s a well thought and researched article questioning the legality and public interest reasons behind the restrictions put by Correios on this dataset:
@herrmann and @codrina
We are updating the GODI evaluation for Brazil this year and this emerged as a dilemma.
I think that if the government provides a dataset that partially qualifies for the GODI (even if not entirely), it should be evaluated in place of the Correios dataset, which is not available.
But the revision process considered that this dataset was absent because 1) Correios should be the publisher of this data and 2) IBGE data doesn’t qualify.
I am not sure of what to do in this situation.
I would like to hear your thoughts on that!
Thank you,
Wagner @Davi_Ventura
I think an important question we have to ponder in the GODI is what is this dataset about. What is the most important thing we care about measuring here? Is it about locations, about addresses or about postal codes? If it’s about locations and addresses, I think the IBGE dataset should probably qualify. If, on the other hand, it’s all about postal codes instead, then only a dataset provided by Correios could be considered an official source, as argued by @codrina.