Using a database of all substance-abuse healthcare providers in the US, teams first explored visualization strategies, then devised ML approaches to provide a “recommendation engine” for end users.
This hackathon was a bit different from the first (toxic comment classification). The data set was a collection of all substance abuse healthcare providers in the US. This data was generated internally by RAND researchers by scraping various text documents. This is the only central repository of its kind. The goal for the hackathon was to “play” with the data from a visual standpoint, to try and explore patterns, trends, gaps, etc through images. The second part of the hackathon was to design a recommendation engine that would enable users to find appropriate local services.