From: "Stian Haklev" <[log in to unmask]>
Date: Wed, 26 Sep 2012 15:15:13 -0400

Subbiah,

I have been thinking for a long time that what is really needed is to
build up a truly free competitor to Google Scholar, similar to what
OpenStreeetMaps is to Google Maps. Google Scholar is wonderful, and I
use it everyday, however it is essentially a black box - there is no
way of finding out which sources it reads from, or to submit edits to
the metadata (I had a class paper I did for a teacher listed as
authored by my teacher with me as the co-author). There is no API, and
the Google Scholar actively resists being accessed by a program (I am
working on a scholarly workflow/citation management program
(http://reganmian.net/wiki/researchr:start), and I would very much
like to do a search in Google Scholar and then do stuff with the
search results - impossible). You cannot download the database (like
you can download Wikipedia) and do citation analysis, etc etc etc...

And of course, much of the data is incomplete because it is based on
heuristic analysis of PDFs and websites. We should at least start with
the Open Access literature - thousands of open access repositories,
and open access journals... if we could ensure that they all output
metadata in a computer readable format (BibTex, BibJSON etc), that
would be a good start. If we also asked all repositories and journals
to ping a server everytime a new publication is added (just like all
blogs already do), we'd already have quite a powerful service...

Stian


On Wed, Sep 26, 2012 at 12:37 AM, Subbiah Arunachalam
<[log in to unmask]> wrote:

> Bang for the buck is one concern. I am more interested in improving Google Scholar so that it becomes inherently better than the fee-based services. I am sure a number of experts who are "open source minded" will be ready to help, but will Google be ready to accept the help and improve this particular service? Long ago I wrote to them but nothing much happened.
>
> Arun