A CSV file that maps bill codes to popular nicknames and keywords.
If you run a service that allows people to search for bills, this would be a great dataset to integrate into your search.
This dataset is designed to be responsive to user searches, so it's not always a literal match between a bill and its keyword. If someone searches for "SOPA", this should match on both SOPA and PIPA, because many people don't know the difference, and the two are often confused or referred to as a single entity.
Similarly, "Obamacare" should match on both the Patient Protection and Affordable Care Act and the associated Health Care and Education Reconciliation Act, which were passed as a package deal and both collectively define what people think of as "Obamacare".
One possible way to have both precision and flexibility would be to add another column that has a relationship flag, to distinguish between terms that refer to a bill exactly, or terms that are less directly related. Open an issue if you want to talk about that.
The CSV has 5 columns.
bill_type - The bill's type. One of:
- "hr" - Bills originating in the House (often seen as "H.R.")
- "hres" - Resolutions originating in the House (often seen as "H.Res")
- "hjres" - Joint resolutions originating in the House (often seen as "H.J.Res")
- "hconres" - Concurrent resolutions originating in the House (often seen as "H.Con.Res")
- "s" - Bills originating in the Senate (often seen as "S.")
- "sres" - Resolutions originating in the Senate (often seen as "S.Res")
- "sjres" - Joint resolutions originating in the Senate (often seen as "S.J.Res")
- "sconres" - Concurrent resolutions originating in the Senate (often seen as "S.Con.Res")
bill_number - The bill's number. (e.g. for "H.R. 3590", this would be "3590")
congress - The number of the Congress this bill was introduced in. (e.g. for the 111th Congress, this would be "111")
term - The term or phrase that should be associated with this bill. Case insensitive, all lower case.
comment - Any comment to describe why this bill should be linked to this keyword.
If you want to add entries, fork it, add away, and send a pull request. I'm inclined to take an inclusive approach, since false positives don't seem as bad as simply missing out on finding the bill someone was searching for altogether.
Public domain information, no restrictions whatsoever.
However, if you split this off somewhere else and add your own expertise to it, please keep it public. Most ideally, keep it within Github, so that your additions can easily be seen and retrieved.
Project initiated by Eric Mill at the Sunlight Foundation.