RE: Critique of OA metric

["Nat Gustafson-Sundell" <n-gustafson-sundell@northwestern.edu>]

<<If you don't believe this, try it.  Create a Web site (it takes 
two minutes) using a blogging service such as WordPress.  Post an 
article to it.  Post another article (or the same one) to an 
established online venue.  The article in the established venue 
will show up higher in search engine rankings.>>

Of course, your experiment only works as such in the first month, 
depending on other factors, though the ranking algorithm will 
include measures of how many like-resources are linked to your 
website, the length of time the website has been around, the 
number of times that specific website has been chosen from among 
results, and so on.  While these kinds of measures strongly favor 
established venues in the short run, there are means for new 
venues to gain attention and to begin appearing in top results, 
usually through indexing services and linking (other shortcut SEO 
measures generally backfire), but best of all by updating 
periodically or frequently with substance on subject.

Ranking algorithms are always evolving and are, of course, 
well-guarded, but I see results changing all the time.  For 
searches by keyword to subject or area (and you can test this 
with new journals that have appeared in the past 5 years), I 
don't think earlier established venues are favored at all, at 
least not in my experience searching for literature.  In fact, I 
choose to use Google sometimes because it will get me outside of 
the box.

I've heard reports that the new library discovery interfaces will 
tend to do a better job of finding good results than the generic 
engines, but haven't had the opportunity to test yet.  Brings up 
an interesting thought however -- in the absence of metadata 
indicating resource-level substance and quality, will the 
collection become more prominent (i.e. since I'm getting this 
using the library's discovery interface or because the Google 
Scholar results show my school branding, I know it is good...).

Next gen library systems are expected to pay a great deal more 
attention to the new variety of electronic resources (aside from 
just articles), to include flexibility in metadata, to include 
recommender services, to push services based on major, etc. and 
so on, so resource consumption could become more context driven 
and less blind.  Stay tuned til next week.

-Nat


-----Original Message-----
From: owner-liblicense-l@lists.yale.edu
[mailto:owner-liblicense-l@lists.yale.edu] On Behalf Of Joseph Esposito
Sent: Thursday, October 29, 2009 5:20 PM
To: liblicense-l@lists.yale.edu
Subject: Re: Critique of OA metric

David's post is an interesting one.  He is certainly correct that
search engines (Google is simply the most prominent, but not the
only means of discovery) "atomize" collections of papers.  This
could lead someone to believe that the "wrapper" of an atomized
article is no longer important, but this overlooks the mattter of
search engine rank.  It is one thing to find an article about
hypertension or a specific aspect of materials science, quite
another for that article to rise to the top of a long list of
potentially relevant Web sites.  For search engine ranking, such
matters as brand are very important, as they collect online
attention and links and lead to higher scoring.

If you don't believe this, try it.  Create a Web site (it takes
two minutes) using a blogging service such as WordPress.  Post an
article to it.  Post another article (or the same one) to an
established online venue.  The article in the established venue
will show up higher in search engine rankings.

As a footnote, I don't believe David is correct in his discussion
of PLOS One.  PLOS One is borrowing the brand of the PLOS
flagship journals.  This is a tricky business.  It works fine
until it doesn't.  Readers are coming to PLOS One (presumably
authors, too) thinking they are getting the editorial rigor of
the PLOS flagships, but they aren't.  Can this go on forever?
Perhaps.  But consider this:  when you buy a telephone handset
for a landline phone that bears the AT&T brand, does it matter
that AT&T has not manufactured handsets in years?  It matters to
me.

Joe Esposito

On Wed, Oct 28, 2009 at 3:54 PM, David Prosser
<david.prosser@bodley.ox.ac.uk> wrote:

> It is simple to conceive of a decoupling between the journal
> 'quality' brand (is the research correct?) and the journal
> 'alerting' brand (here is a group of articles that may be of
> interest to you) as described by Sally.  In the print world
> grouping papers thematically together made perfect sense; in
> the online world where people increasing use search engines
> (whether specific, like Medline, or general, like Google) to
> find papers it is perhaps less useful.
>
> So, a journal table of contents e-mail may be useful, but
> equally I may be more interested in seeing the daily digest of
> papers with a particular tag in Connotea, say. That way the
> community would define its own interests rather than having the
> collection codified by an editor.  And different communities
> could combine the content of different journals in different
> ways.  This to me is what PLoS One has done - provide the
> quality brand, but leave the 'what's this journal about' to the
> readers.  It seems inevitable (and was once titles and
> abstracts went online!) that the 'alerting' brand is going to
> become less and less within the control of publishers and more
> in the control of users.
>
> David
>
> David Prosser
> SPARC Europe