On Thu, Sep 25, 2008 at 12:53:49PM +0100, Stuart Lewis wrote:
> On 25/09/2008 12:40, "Talat Chaudhri" <[log in to unmask]> wrote:
> We could in theory make repositories see that a request for the full text
> has come from a Google search and hijack this and instead return the
> metadata page.
>
> HOWEVER... when the search engine spiders find you doing this, they decrease
> your ranking as they see that you are 'cheating' because you return
> different web pages to their spiders as you do to end users for an identical
> request.
>
> AFAIK there is no clean technical way to force people to the metadata
> jump-off page.
We've had to deal with this question at arXiv. As soon as Google
started to harvest our PDF we saw a huge shift in the PDF/splash page
download ratio with lots of people coming straight to the PDF.
First, I think one has to come to terms with the fact that many users
may _prefer_ to go straight to the PDF. Who are we to tell them they
can't? (Even if we put lots of effort into that splash page.) Back in
the day download times meant it was worth reading the splash/abstract
first but that typically doesn't apply now.
Second, at arXiv we add a "stamp" to the left hand margin of the first
page of (most of *) our PDF files. These are hyperlinked to the splash
page to provide a way to get that from the PDF. See for example:
http://arxiv.org/pdf/hep-ph/0010338v2
Cheers,
Simeon
* There are some technical issues which mean we can't stamp a few of
our articles.
|