Comments on: Map of Science

Map of Science

homunculus — Sat, 21 Mar 2009 20:55:29 -0800

Knowledge, in Real Time. "A new picture of science — and possibly future innovation — comes into focus with the mapping of scientists' online research behavior."

By: a non e mouse

a non e mouse — Sat, 21 Mar 2009 21:09:09 -0800

Excellent - thanks for this. Finally, a good use of clickstream data.

By: homunculus

homunculus — Sat, 21 Mar 2009 21:10:13 -0800

A New Picture of the Two Cultures

By: homunculus

homunculus — Sat, 21 Mar 2009 21:11:23 -0800

By: Baby_Balrog

Baby_Balrog — Sat, 21 Mar 2009 21:58:18 -0800

*bong hit* I'm especially interested in the line connecting analytical chemistry to religion. *exhale*

By: delmoi

delmoi — Sat, 21 Mar 2009 22:01:23 -0800

Those 'constellation' style visualizations of large graphs aren't really that useful. A straight up adjacency matrix visualization can be easier to read, I think.

By: Cold Lurkey

Cold Lurkey — Sat, 21 Mar 2009 22:09:59 -0800

Okay great, but citation data, like clickstream data, is useless, from an actual scientific standpoint, without context. I could cite a paper endlessly even if why i'm doing it is disproving everything that they claimed. i.e. while previously reports found x(endnote cite) we found by more carefully, more rigorously, generally checking our shit out better, y. Likewise, I could download a paper because they have the right idea of a technique, but the wrong execution and I would never cite them for that. Second, so what? I'm not getting the real take home message here... is this just data for data's sake or does it tell us something else we didn't know about the nature of scientific inquiry

By: Pinback

Pinback — Sat, 21 Mar 2009 22:49:09 -0800

It's interesting (I say, as I sit here browsing papers to choose for a short write-up on the usefulness/applicability of biological control techniques), but I wonder if it's not more indicative of random browsing and mind-wandering as much as anything else. For instance, I get the strong flow of connections between the ecological/environmental/biological sciences, right through to the chemical & physical sciences, but there's also an almost-equally strong connection between the eco/enviro/bio sciences & the arts/social sciences? That suggests that it's mapping general interests of particular groups (i.e. enviro/bio types look to have a side interest in architecture & similar) as much as research connections between disciplines. And the poor geo's stuck out there by themselves. That'll go down well with some of my friends ;-)

By: Aquaman

Aquaman — Sun, 22 Mar 2009 00:08:02 -0800

It would be even awesomer to see an animation of this graphic showing change in scientific activity focus over time. That's right, I said awesomer. Deal with it.

By: hal_c_on

hal_c_on — Sun, 22 Mar 2009 03:47:18 -0800

I don't know much about modeling, but I would put a LOT of money on the idea that this would be much more relevant and useful if it this "map" were to follow natural maps...like a fractal. As of now...this just looks douchebaggy since its kinda closed. Weak, los alamos...weak.

By: srboisvert

srboisvert — Sun, 22 Mar 2009 09:18:51 -0800

By: Mental Wimp

Mental Wimp — Sun, 22 Mar 2009 12:04:36 -0800

Curious. No reference to mathematics beyond the little group labeled "Statistics"? Or did I miss it?

By: a non e mouse

a non e mouse — Sun, 22 Mar 2009 14:31:31 -0800

economics is teh maths

By: Maias

Maias — Sun, 22 Mar 2009 14:45:13 -0800

If you could in real time see what your competitors are searching, that could be a serious privacy/etc. issue. Because although the guy who beat you to the last discovery might be searching for a particular paper to disprove it, he might also have the same idea you have now and if you know what he's searching, you could potentially get a pretty good clue about what he's thinking. I guess this doesn't work on that fine-grained a level though...

By: Aquaman

Aquaman — Sun, 22 Mar 2009 19:28:49 -0800

Yeah, I was thinking it would be more like a slow ballet of shifting intensities of scientific interest, rather than a one-to-one map. Like right now acoustics is pretty quiet, but as soon as the iCochlear Implant is developed, I bet there would be a lot more links developed between it and cybernetics, for example. Hey look, a that squirrel has a fluffy tail!