I was really surprised to see how little research is done in the CS fields until I realized something: the lines only show the quantity of citations going in and out across different fields.
They have bio and med listed separately when they're really closely tied - so they naturally have a huge amount of information exchange going on between the two, making for the seemingly-gargantuan research count/amonut for those areas.
In CS, on the other hand, the amount of research is huge (probably biggest of all since the barrier of entry is so low and isn't limited to people of that field alone); but it's mainly CS people citing other CS people's work(s). We effectively live in a bubble, albeit a very huge one. We interact and share with one another, but not that much with the rest of the industries - and we probably should, since there's a lot of ideas in cyberspace that can be mapped directly to real life.....
The bubbles represent the size and importance of the field. This is an accurate depiction of what science is like today.
Science and Nature are filled with articles about medicine and biology and chemistry and physics and so on. These are what scientists think of as science. Rarely does a math or computer science paper slip in, except in the case of groundbreaking new work. And yet most CS professors would give anything to have a paper of theirs published in Science. So you can see how the rest of science values our contributions: every once in a while we contribute something interesting, but for the most part we are not noteworthy. And this is the way it should be. Bipartite graph algorithms and topological manifolds are not as important as Gardasil and alternative energy.
A substantial number of doctors doing useful research these days are, surprise, Medical Doctors. If you want an eye opening experience, take a trip over to the nearest medical library and spend some time going through the journals. It will radically change your opinion of what science is and how your field fits into it. Discrete Math is a pretty cool journal, but it's kind of embarrassing to hold it up to The New England Journal of Medicine. Even Annals of Math is, well, pretty irrelevant.
There's a long tradition of specialists taking myopic views of the world. Computer Scientists think the world is one big computer (http://focus.aps.org/story/v9/st27), chemists think life is one big chemical reaction, stone-cutters think everything can be explained in terms of stone-cutting (The Stone Diaries). Rarely is that the whole truth. You have to be careful to step back and take your own bias into account.
Bipartite graph algorithms and topological manifolds are not as important as Gardasil and alternative energy.
I highly disagree. Drop by any R&D lab at your closest university and see how badly they depend on the latest software, the most efficient algorithms, and other CS-derived "sciences."
If it weren't for languages like Matlab & the libraries of numerical methods and algorithms, alternative energy R&D wouldn't exist. I was at the Nuclear Engineering department of my university earlier today, and they're using CS stuff we haven't even looked at yet!
CS is far more important in the grand scheme of things than it's being given credit for, but we silly hackers are content to appreciate ourselves and not look for more.
Why is physics so big, do you think? We're not going to be using TeV physics for a long time yet. Is it just coasting on last century's reputation? That graph puts material engineering on a separate dot.
Their site shows a similar, higher resolution map of science and maps it to a sphere. Pretty interesting.
Running a literature search, Richard Klavans and Kevin Boyack published papers concerning science mapping between 2005-2006, so I imagine there was a good chance they were inspired by eigenfactor.
Tamara Munzner was doing this with hyperbolic trees back in the mid-90s at Stanford. I remember this because I always thought it was the most sensible interface for big graphs.
Perhaps computer science is the glue rather than the cutting edge?
A lot of research is not even realistically possible without computers.
Parallellisation is important for bioinformatics and so on.
But the really cool stuff that computers could potentially solve is incredibly hard, like machine translation and computer vision.
But if the singularists are right computer science will solve the problem that solves all other problems :)
Or destroys us :(
This article brings up an idea I had for a while: make a map of the sciences (and eventually of all academic disciplines) with meaningful x and y axes. My suggestion is practical vs. theoretical for one dimension, and hard vs. soft for the other. Has this been done before?
They have bio and med listed separately when they're really closely tied - so they naturally have a huge amount of information exchange going on between the two, making for the seemingly-gargantuan research count/amonut for those areas.
In CS, on the other hand, the amount of research is huge (probably biggest of all since the barrier of entry is so low and isn't limited to people of that field alone); but it's mainly CS people citing other CS people's work(s). We effectively live in a bubble, albeit a very huge one. We interact and share with one another, but not that much with the rest of the industries - and we probably should, since there's a lot of ideas in cyberspace that can be mapped directly to real life.....