OA Growth Monitoring Needs a Google Data-Mining Exemption

[C. Sean Burns]

On 08/23/2013 11:13 AM, Stevan Harnad wrote:
> Yes, hand-sampling can and does provide valuable information.

> But, as I said, for systematic ongoing monitoring of the global
> time-course of OA growth across institutions, disciplines and
> nations, hand-sampling is excruciatingly difficult and
> time-consuming,

The key point in sampling is that it can be cost effective, in the 
relevant ways, for capturing, analyzing, and inferring from "small" (cf. 
to 20,000 items) data sets. If a doctoral student can do it, with some 
degree of statistical confidence, then .... Besides, it's not all that 
difficult to write programming scripts that can do some heavy work.

> holding research that could greatly benefit the worldwide research
> community (as well as Google and Google Scholar) to a scale and pace
> that is more suitable for a doctoral dissertation.

Hmmm.

> Historically speaking, if a few projects designed to monitor the
> ongoing global growth and distribution of OA were allowed to do
> machine data-mining in Google space, the growth rate of OA would be
> dramatically accelerated (and thereby also the size and functionality
> of Google Scholar space).

It's not clear how such monitoring and how the release of such 
information functions as a causal factor for accelerating OA. Besides, 
who does Google grant such a privilege?

> Otherwise, efforts to enrich Google Scholar space are relegated to
> the same fate as attempts to enrich vendors, spammers, napsters or
> phishermen.

It's also not clear, in this example, how Google Scholar is obligated or 
has the incentive to treat researchers any differently than those you 
mention. Especially since it's not all that clear how Google benefits 
from Google Scholar, as a number of people have wondered.

Yours,
Sean Burns