3 ms·
That's true, the priority of U.S. News and World Report reflects not just the popularity of that particular list, but that most of those schools have fairly can
by danso 10y ago
That's true, the priority of U.S. News and World Report reflects not just the popularity of that particular list, but that most of those schools have fairly canonical names that are easy to text-mine against ("text-mine" in the most crudest form, string matching and some regexing)...this project started out as an attempt to find out how many Congressmembers had reached various levels of higher education (e.g. B.A.,B.S.,MBA, etc) but the descriptions of educational events is too inconsistent in the Congress bioguide (that's even before considering how school names changed over the centuries).
So I decided to make the text-mining easier by just searching for well-known schools. It's possible that a few of the big public prestigious schools, such as University of Michigan, would be near the top. The raw text from each Congressmember's biography can be found in the repo:
https://github.com/dannguyen/congress-colleges/tree/master/data/raw/bioguide/text https://github.com/dannguyen/congress-colleges/tree/master/d...
It's probably possible to (easily) automate the classification of how many Congressmembers served in the military or passed the bar...but I can save that for another day.
- arcanus 10y agoIt is an interesting and cool effort! I would however at least provide the option to sort by congressmembers, and likely would make that the default. That would be more informative: anyone can look up the USNEWS Top-25.
- showerst 10y agoVery cool! Would NNDB maybe have more normalized data? http://www.nndb.com/people/106/000122737/ http://www.nndb.com/people/106/000122737/
- pash 10y agoInstead of mining the bio guide, I would mine Wikipedia. There is a page listing all members of the current U.S. Congress [0], and that page links to each member's own Wikipedia page. Almost without exception, each current member's personal page names the universities the member graduated from in the info box. So mine the membership list for the members' names and the URLs of their personal Wikipedia pages, then mine the info box on each personal page for university affiliations. The links will allow you to disambiguate universities with similar names. Pages also exist listing the membership of almost all former U.S. Congresses, so you could repeat the process going back in time; it would be interesting to see how affiliations have changed and diversified (or not) over the years. A similar process would work for finding the academic affiliations of members of the U.S. Supreme Court and the president's cabinet, as well as of state governors; it might even work to find comparable information for some sets of foreign poliicians (e.g., British MPs or members of European Parliament). (And I agree with arcanus; you should change your page to sort by the count of affiliated members of Congress, not U.S. News rank, as it's the former that is the information being presented, not the latter.) 0. https://en.m.wikipedia.org/wiki/114th_United_States_Congress https://en.m.wikipedia.org/wiki/114th_United_States_Congress
- danso 10y agoI haven't looked at extreme edge cases, but you're right, Wikipedia tends to have the right alma mater, included situations in which someone has only attended classes but isn't reported as actually matriculating. It wouldn't require a big traversal of Wikipedia (though I imagine that would be easy enough)...the unitedstates Github repo, which is a semi-automated crowdsourced repo of Congress data, has all the historical Congressmembers as structured YAML, including their Wikipedia page (though I haven't counted to see how many have that attribute filled): https://github.com/unitedstates/congress-legislators https://github.com/unitedstates/congress-legislators
- ZeroGravitas 10y agoI was going to suggest looking at http://everypolitician.org http://everypolitician.org but it seems like the US is ahead of the game in releasing this info. People from other countries that want to do similar might find it useful though.