2.1 The original story
When people write FOIA requests, it can be pretty stressful to wonder if they’re going to get fulfilled. Unless you’ve done it many times before, the process might feel like a little bit of a black box. what exactly do you need to say in order your request filled?
Enter the FOIA Predictor! You insert your FOIA request, you say who you’re sending it to, and it gives you an estimate of what your chances look like.
The FOIA Predictor is an excellent case study in a lot of things that do with machine learning:
- classification problems
- feature selection
- algorithm choice
- evaluation metrics
- explainability
- community development
- and a whole lot more!
FOIA Predictor was written by Rachel Downs when she was an intern at data.world, which we’re going to use to explain away any of the rough edges the FOIA Predictor might have. You can’t hold anything an intern does against them, because: they’re an intern!