Tara Calishain

I dream in data flows

Hooking the ProPublica Congress API and the FCC Licensing API together and pointing them at Google

This recent political stuff has impressed upon me that I should be doing more to connect politician data to local news data. Furthermore, as long as I’m doing that I should be making news monitoring tools. So that’s what I’ve been doodling around with today. Here’s how it works – currently I’m working on just the House of Reps.

You enter a state and a congressional district number (I do not have the $$$ to afford an address-to-district API) and the following things happen: the ProPublica API is called to get information on the representative for that district. At the same time, the FCC API is called to get the TV stations in both the capitol of that state and the largest city in that state (they are often not the same.)

The websites for the TV stations are aggregated into Google searches along with the name of the representative. In other words, you end up with Google searches for the name of the rep in the Web spaces of the TV stations in the rep’s state capital and largest city. The searches come in two flavors: general, and last 24 hours only.

That’ll be useful and I’m sure I’ll have fun playing with it, but I want to go further. An obvious thing to do is make the Google search generated available as an RSS feed, so I’ll do that. I might also try to make a nationwide search for the rep’s name and state name as an RSS feed via Bing News and its loc: syntax, but I’m not sure about that yet.

Another thing is to add the option to break your searches out by Congress session date, For example, I could set up an option that you could search through the timespan of the 117th congress. I’m not sure how far back that would be useful — maybe the 110th?

The ProPublica API includes a lot of information for each rep, including social media accounts, so I thought another fun option would be to assemble Wayback Machine archive URLs as well as site: inurl: searches on Google. Maybe have time options for those by congressional session too just to be weird

If I want to get really stupid I’ll integrate StreetScoop ( https://searchgizmos.com/streetscoop/ ), which takes an American address, finds a nearby TV station, and searches their news archives for that street name. Enter the rep’s local office address and get a Google News search for that street from a relevant TV station. And then save it as an RSS feed because why not.

There are so many ways to identify people via the constant metadata of time and space. When you combine that metadata (like space) with authoritative sources to use that metadata (like FCC license databases) you provide both a focused data pool to search in and a data pool less likely to have misinformation and other infosewage.