Extracting Data on Journals

From Eureka

The current math journal data has been extracted from the 2002 AMS Journal Price Survey.

The following sites contain excellent information on academic journals (not just mathematics ones), from which it should easily be possible to extract data:

• www.journalprices.com This site has been setup by two academic economists, Ted Bergstrom and R. Preston McAfee. They have been very active in the whole issue of overpriced journals (see this page). It uses 2004-2006 data obtained from Ulrich's International Periodical Directory and ISI Journal Citation Reports (specifically the JCR dataset, I'm not quite sure what that means). This data is available in an Excel format. (my attempted download of this 1.4Mb file aborted for some reason. Does it work for anyone else?) It can also be downloaded as a comma-delimited text file (click on Mathematics and Format results as tab-delimited text to get a list of 401 math journals). It is an excellent resource, and and a lot of effort has gone into the numbers they produce. They have already been informed about this wiki, and are very co-operative.

Update: Reading through the explanation given by Rob Kirby and Colin Rourke for why Geom. Topol. and ATG Algebr. Geom. Topol. are no longer free underlines the motivation for this site. It also brings up some interesting points : Geom. Topol. and ATG Algebr. Geom. Topol. can't be found at www.journalprices.com or www.eigenfactor.org. On the other hand they can be found at Journal Info, but their eigenfactor data does not display there. This site can't use the JCR Impact Factor since it is not free.

Update: For anyone who wants to play with it, journalprices.com data on 923 math and physics journals is available in tab-delimited format here.

• www.eigenfactor.org This site has been setup by the Bergstrom lab in the Department of Biology at the University of Washington. It is another excellent (and new) resource. It also uses the Thompson Scientific JCR dataset. It calculates the "Eigenfactor" and "Article Influence" of a journal in a manner similar to Google's PageRank algorithm. The "Article Influence" measures roughly the same thing as JCR's "Impact Factor"... but is more advanced. Even better, it's free! It also meshes with the journal pricing data from www.journalprices.com.

• Journal Info, run by Lund University in Sweden. This combines the information from the above two sites with some extra data, such as the Sherpa data which indicates whether authors are allowed by the publisher to self-archive their papers.

• One can also obtain citation data from MathSciNet. This data is not free; although perhaps an agreement could be made with the AMS. For instance, for journals where eigenfactors are not known (such as Geom. Topol.), perhaps the AMS would allow wiki contributors to manually add AMS citation data in this way.

The first three sites above are all free, and all interlinked, sort of. But it should be a piece of cake to get the data out, especially since all of them are freely available resources run by people in the academic community.

Thus we can update our pages on each journal to have the data stored above : Name, Publisher, Price, citation ranking, resultant value scores, for-profit/for-free. I suggest we just get the maths journals so long, until physicists start to become interested in the site.

Then we need to start providing extra content, like editorial board, average time between submission/acceptance, etc., and grouping the journals into subjects (Topology, Functional analysis, etc.).

• The Directory of Open Access Journals lists 125 in math and statistics and 54 in physics and astronomy. Getting the information out might be a little tricky. Blake Stacey 16:59, 27 August 2007 (EDT)

Some of the journals listed at DOAJ strike me as being not-very-famous, for instance "Matemáticas : Enseñanza Universitaria". On the other hand, is there a recognizable consensus about what is a "mainstream" journal and what isn't? I don't think there is. --Brucebartlett 20:28, 27 August 2007 (EDT)

I would argue that "indexed by MathScinet" is a pretty good measure of "mainstream" (though they go pretty obscure. I recently did a MathSciNet review for a paper published in the Bulgarian Journal of Physics). --Ben Webster 14:46, 28 August 2007 (EDT)