GovTrack Wiki

From GovTrack.us Wiki

Here are some possible improvements to GovTrack that you, as a coder, might be able to work on:

Contents 1 Scraping more data 2 Parsing bill text 3 Data analysis 4 Fixing site problems 5 Improving site features 5.1 Bills: 5.2 Representatives 5.3 Miscellaneous 6 Extending the site to new areas 7 Notes

Scraping more data

• Committee and conference reports, and tying them into the relevant bills
• Finding earmark information out of these reports, though it seems to often be embedded in images (OCR?)
• Scouring for news articles relevant to legislation
• Scraping some general information about committees, and organizing some historical data
• See what's on speaker.gov and in the Weekly Leader for tracking legislation about to get activity
• CBO report fetching keeps breaking. Talk to CBO about getting a machine-processable list of all reports in a stable format.
• OMB-SAP report fetching is probably broken.
• Bug report: <If a member appears more than once on one of your pages from the Congressional Record, your feed excerpts the beginning of the last time that person appears on that page instead of the first. So, instead of getting the introduction to what the members have to say, you often end up with just something like, "I yield the floor" which is not particularly helpful.>
• Bill actions in committees may not be ordered properly. See email from DRM to JT 3/20/07.
• Bill actions aren't labeled with whether they occur in the House or Senate, but I think it is scrape-able.
• Congressional Record parsing bugs, see http://www.govtrack.us/congress/record.xpd?id=110-h20070306-79
• Nominations
• Personal financial disclosures, and see if we can re-open discussion with LegiStorm [1]
• Election results: http://www.fec.gov/pubrec/fe2006/federalelections2006.shtml How safe a district is each rep in?

Parsing bill text

• Finding earmarks
• Relating bills to the laws being amended: showing how the law would change, putting changes in context. Showing popular names of titles being modified, show the bill that originated the law.
• Comparing bills --- e.g. find where one bill is incorporated into another, shared provisions, House and Senate companion bills. Simhash might work (though I tried once without success) or a similar approach, see plagiarism software. A user writes <If you want a couple bills to practice on as you test your implementation, here's an article http://public.cq.com/docs/cqt/news110-000002520575.html ...about how this bill... http://www.govtrack.us/congress/bill.xpd?bill=s110-1537 was added to this bill: http://www.govtrack.us/congress/bill.xpd?bill=h110-2206 And here's an article about the majority planning more of the same: http://public.cq.com/docs/cqt/news110-000002499096.html>
• Tracking the evolution of the text of bills better: Comparing a bill to itself over time
• Parse references to public laws (e.g. http://www.govtrack.us/congress/billtext.xpd?bill=h110-995) and hyperlink them on the site

Data analysis

• Visualizations of legislation language and evolution
• Visualization of legislative statistics, generation of new statistics and charts, finding oddities in the data
• An automated way to summarize a legislator's performance, including what issues he's worked on
• Showing some earmark related statistics: total $$$$ allocated

Fixing site problems

• CBO issues updated reports on bills. We need to be able to store multiple reports per bill and get updates as appropriate.
• Making sure OpenID registration and log in actually works
• DONE (JT) Improving the look and usefulness of the embeddable widgets: Let users choose colors, etc.
• Clean up the committee pages, make it easier to sift through legislation and navigate members
• Make the website fully compliant with accessibility standards
• Get images for all members of congress. Historical images would be nice too. Scan old publications of the Congressional Directory if it's not on bioguide! They're in libraries.
• Improve site responsiveness (and decrease load on server): Hold some more things in memory? Improve Mono's XSLT engine? Simplify XSLT templates? Different handling of feeds?
• Remove all uses of the old "date" attribute (Perl time, seconds since 1/1/1970) in the various data XML files and replace them with uses of the newer "datetime" attribute (in the readable ISO format).

Improving site features

Bills:

• Linking to (or data integration?) bill wiki websites- washingtonwatch, congresspedia, readablelaws.
• Linking to (or data integration of?) MAPLight.org on bill and people pages
• Search/browse legislation and amendments by sponsor
• Tagging bills (topic areas, which bills are must-pass)
• Show on bill pages which parts of the US Code it modifies
• Show on bill pages a more graphical representation of how a bill becomes a law (www.senate.gov/reference/resources/pdf/legprocessflowchart.pdf, http://www.lexisnexis.com/help/cu/The_Legislative_Process/How_a_Bill_Becomes_Law.htm)
• We have detailed action information but we don't display it. Each action also may have a reference to the congressional record, so we should do good linking on that as well
• Can we search by public/private law number? (Maybe we can already.) Make a list of all public laws.
• In list of roll call votes, include votes on rules related to the bill (see http://www.govtrack.us/congress/votes.xpd?bill=h110-2669, http://www.govtrack.us/congress/votes.xpd?bill=hr110-637)
• Status of legislation table like http://thomas.loc.gov/home/approp/app08.html
• Do something throughout to clarify what "Rule providing for the consideration of..." bills actually are.
• Tell people why they should link to bills on GovTrack: use the status widget, permalink, can link to paragraphs in bill text...

Representatives

• Zoom into congressional district by zip code
• A table of how a user's representatives voted not just by issue but also by tracked bills
• Integrate feeds from members of congress: blogs, even twitter (http://blog.sunlightfoundation.com/2008/06/11/following-johnculberson/ http://www.techpresident.com/blog/entry/26342/congresstwitters)!!

Miscellaneous

• Our own Daily Digest type of page (see the Daily Digest on THOMAS).
• Better navigation of Congressional Record. Navigate from section to section. More hyperlinking. Integration with video.
• Show on the votes pages the time the vote took place (do we have this data?)
• Forums or other mediums for discussion--- but constructive discussion, not rants and opinion-shouting
• Votes page for a particular bill should link back to the bill status page.
• Procedural actions need explanations. On vote pages, this is the vote type or question that is displayed. On bill pages, this is the last action (or if we show more actions, then those).

Extending the site to new areas

• The Constitution
• U.S. Code: Show how the code would be amended by legislation not yet enacted (http://www.statutelaw.gov.uk/Home.aspx)
• Federal regulations
• Judicial documents
• State-level legislation, etc.

Notes

I (Josh) am more than happy to help anyone work on these things if you're interested.