Helping people find and understand state-level bills is incredibly hard.
First, not every bill is interesting to every person. A bill on education might be interesting to teachers, but not the general population.
Second, most bills are a nightmarish jumble of legal jargon. How can ordinary people understand these bills?
We solve the problem of finding bills using FastText, an open source AI classifier. National level bills are categorized into a useful topic hierarchy, so we train the classifier on 100,000+ national level bills. Then, it can automatically categorize state level bills.
Check out the nlp folder to see the code!