The Serenata de Amor Operation arose from a combination of needs, from many people, of many nationalities:
- seeing Machine Learning being better applied to improve population's lives;
- learning whom to vote;
- doing something about the corruption problem present in the whole world.
We are building an intelligence capable of analyzing public spending and saying, with reliability, the possibility of each receipt being unlawful. This information will be used beyond the code, in the world outside of GitHub. Everything is open source from the beginning, allowing others to fork the project when their ideas diverge from the Operation Serenata de Amor.
Our current milestone is to create the means for this kind of automation with the Quota for Exercising Parliamentary Activity (CEAP), from the Brazilian Chamber of Deputies. This job includes the development of APIs, data cleaning and analyses, conception and validation of scientific hypothesis, confirmation of illicit acts via investigation and reports - to the population and to legal authorities.
To achieve this goal, unprecedented, we invite everyone to train the intelligence, collect information, cross databases, validate hypothesis and apply Machine Learning with models competing against each other and getting combined in ensembles with higher precision than any previous option.
Visit out official website and remember to read the contribution guide if you plan to help on the software side.
The project Operation Serenata de Amor is available under MIT license.