Improving Government Response to Citizen Requests Online

Fellows: Garren Gaut, Andrea Navarrete, Laila Wahedi
Data Science Mentor(s): Adolfo De Unanue, Eric Potash
Project Manager: Paul van der Boor
Project Partner: Office of the President of Mexico

Since the the enactment of the Constitution of 1917, all Mexican citizens have the right to petition members of the government and receive an official response. These requests range from serious appeals for services to frivolous solicitations, such as inviting the President to a birthday party. But all petitions, regardless of importance or means of submission, are manually reviewed and responded to, an often inefficient and slow process to deal with over a thousand requests a month.

DSSG worked with the Office of the President of Mexico to improve this system as part of their Estrategia Digital Nacional (National Digital Strategy). We created new algorithms to classify petitions based on their content and importance, route them to the correct government agency, and partially automate the responses to common requests. The project also used machine learning on over 27,000 anonymized digital petitions and gob.mx historic analytics to reveal patterns in the data. These insights can help the Government of Mexico understand community needs and provide targeted services, faster responses, and reduce information asymmetry between government and citizens.