![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Science is Friggin' AWESOME!
Jun. 14th, 2013 09:57 amFound a whole bunch of geeky info over the last couple of days:
Depending on your viewpoint, this is either absolutely fascinating or completely horrific, or both simultaneously. Me, I'm still snickering over "EDWARDx talk." A Duke professor explains social network analysis, metadata, and big data via Using Metadata to Find Paul Revere: http://kieranhealy.org/blog/archives/2013/06/09/using-metadata-to-find-paul-revere/
Followup: http://kieranhealy.org/blog/archives/2013/06/11/following-up-on-paul-revere/
A few people emailed to say that contemporary SNA* methods differ in significant respects or, more substantively, that network theorists have moved away from the strong picture of pure structure that they advocated in the early days. In response I can only say, again, that the the purpose of the post was not to do a contemporary piece of SNA on Fischer’s data. ... I wanted to give non-specialists a sense of how the structural analysis of what’s being called “metadata” works, and to show in a fun but hopefully telling way how much you can get out of that approach.
*SNA = Social Network Analysis
As a bonus, here's a bit from "Bang Goes the Theory" - which, as far as I can tell, is proof that the British-American "let's steal TV show ideas" doesn't flow entirely States-wards. (You'd think that a British knockoff of Mythbusters with "bang" in the title would have more explosions, wouldn't you?)
But I nitpick. Bang Goes the Theory brought in a mathematician to finally explain the math behind population estimates in a way that I could grasp, and it was cool enough that I'll throw that here too. The illustration sin the show first used ping pong balls, then applied the same thing to London taxi cabs.
1) Establish your population: i.e., pour a boatload of ping pong balls into a fish tank.
2) Take out 100 and mark them. For the taxi cabs, they were noting license plate #s
3) Release them back into the wild. The ping pong balls got thoroughly stirred; the taxis did their own running around.
4) Pull another 100 as a sample.
5) Find X by noting note how many in Sample 2 had been part of sample 1. For the ping pong balls, x =17.
6) Find Y by dividing the sample size (100) by X. Ping pong balls: 100/17 = 5.88
7) Sample size multiplied by Y = estimate of population size. 100x5.88 = 588 ping pong balls
In both the ping pong balls and the taxis it was more of a general estimate than an accurate one; in both cases he didn't achieve the exact number, but he certainly achieved a fairly close one.
And as a bonus bonus - X-rays have been used to rescue and recreate a lost aria - the composer basically blacked out the entire page with carbon after bad reviews. They've pinged on the iron in the ink to essentially dot-matrix back the words and notes. http://www.wired.co.uk/news/archive/2013-06/11/opera-recovered-with-x-rays
Science is AWESOME, y'all!
Depending on your viewpoint, this is either absolutely fascinating or completely horrific, or both simultaneously. Me, I'm still snickering over "EDWARDx talk." A Duke professor explains social network analysis, metadata, and big data via Using Metadata to Find Paul Revere: http://kieranhealy.org/blog/archives/2013/06/09/using-metadata-to-find-paul-revere/
Followup: http://kieranhealy.org/blog/archives/2013/06/11/following-up-on-paul-revere/
A few people emailed to say that contemporary SNA* methods differ in significant respects or, more substantively, that network theorists have moved away from the strong picture of pure structure that they advocated in the early days. In response I can only say, again, that the the purpose of the post was not to do a contemporary piece of SNA on Fischer’s data. ... I wanted to give non-specialists a sense of how the structural analysis of what’s being called “metadata” works, and to show in a fun but hopefully telling way how much you can get out of that approach.
*SNA = Social Network Analysis
As a bonus, here's a bit from "Bang Goes the Theory" - which, as far as I can tell, is proof that the British-American "let's steal TV show ideas" doesn't flow entirely States-wards. (You'd think that a British knockoff of Mythbusters with "bang" in the title would have more explosions, wouldn't you?)
But I nitpick. Bang Goes the Theory brought in a mathematician to finally explain the math behind population estimates in a way that I could grasp, and it was cool enough that I'll throw that here too. The illustration sin the show first used ping pong balls, then applied the same thing to London taxi cabs.
1) Establish your population: i.e., pour a boatload of ping pong balls into a fish tank.
2) Take out 100 and mark them. For the taxi cabs, they were noting license plate #s
3) Release them back into the wild. The ping pong balls got thoroughly stirred; the taxis did their own running around.
4) Pull another 100 as a sample.
5) Find X by noting note how many in Sample 2 had been part of sample 1. For the ping pong balls, x =17.
6) Find Y by dividing the sample size (100) by X. Ping pong balls: 100/17 = 5.88
7) Sample size multiplied by Y = estimate of population size. 100x5.88 = 588 ping pong balls
In both the ping pong balls and the taxis it was more of a general estimate than an accurate one; in both cases he didn't achieve the exact number, but he certainly achieved a fairly close one.
And as a bonus bonus - X-rays have been used to rescue and recreate a lost aria - the composer basically blacked out the entire page with carbon after bad reviews. They've pinged on the iron in the ink to essentially dot-matrix back the words and notes. http://www.wired.co.uk/news/archive/2013-06/11/opera-recovered-with-x-rays
Science is AWESOME, y'all!
