[Esip-preserve] Esip-preserve Digest, Vol 22, Issue 6

[Bruce Barkstrom]

The original reference for this concern is
http://www.gisdevelopment.net/technology/ip/fio_2.htm and
the page that follows (fio_3.htm).  The issue is not that
proprietary details can be hidden.  Rather, the problem is
that when doing scientific work with data, proprietary processes
get in the way of replicating results.  The reference notes that
IKONOS imagery hides details of the optics that enters the
orthorectification algorithms so that these details could not
be checked.  From my perspective, this means that IKONOS
imagery should not be used for scientific work because it does
not appear one could replicate results obtained from these
data using physical understanding of the instrument.  A specific
concern would be how much radiometric uncertainty the orthorectification
process introduces.  An accurate calculation will require the point spread
function for each pixel - and it is precisely this detail that gets hidden
by the approach identified in the reference.  I suspect that the uncertainty
in the radiometry increases as pixels are further from the center of the
image - although there's no way to check this that I can see.

As an additional note, I would not normally use the term "provenance"
for algorithm derivations or source code.  See my paper in ESI a few months
ago.

Bruce B.

On Tue, Nov 30, 2010 at 1:13 PM, Myers, Jim <MYERSJ4 at rpi.edu> wrote:

> The W3C Provenance exploratory group
> (http://www.w3.org/2005/Incubator/prov/ ) has a business use case along
> these lines - where one might want to provide enough provenance detail
> to show that contractual terms were met without revealing all of the
> provenance known. In the case here, that might be provenance showing
> that the algorithm has been shown to work well on standard test cases
> while not revealing the actual code, etc.
>
> Cheers,
>  Jim
>
>
> James D. Myers, Ph.D.
> Director, Computational Center for Nanotechnology Innovation (CCNI)
> Rensselaer Polytechnic Institute
> 405 Jordan Rd
> Troy, NY 12180
> 518-276-2858
> myersj4 at rpi.edu
>
>
>
> -----Original Message-----
> From: esip-preserve-bounces at lists.esipfed.org
> [mailto:esip-preserve-bounces at lists.esipfed.org] On Behalf Of
> esip-preserve-request at lists.esipfed.org
> Sent: Tuesday, November 30, 2010 12:01 PM
> To: esip-preserve at lists.esipfed.org
> Subject: Esip-preserve Digest, Vol 22, Issue 6
>
> Send Esip-preserve mailing list submissions to
>        esip-preserve at lists.esipfed.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>        http://www.lists.esipfed.org/mailman/listinfo/esip-preserve
> or, via email, send a message with subject or body 'help' to
>        esip-preserve-request at lists.esipfed.org
>
> You can reach the person managing the list at
>        esip-preserve-owner at lists.esipfed.org
>
> When replying, please edit your Subject line so it is more specific than
> "Re: Contents of Esip-preserve digest..."
>
>
> Today's Topics:
>
>   1. Provenance for Data Created With Proprietary      Software
>      (Bruce Barkstrom)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 30 Nov 2010 08:26:02 -0500
> From: Bruce Barkstrom <brbarkstrom at gmail.com>
> To: ESIP Preservation cluster cluster <esip-preserve at rtpnet.org>
> Subject: [Esip-preserve] Provenance for Data Created With Proprietary
>        Software
> Message-ID:
>        <AANLkTi=y79FKpS3LnNnpiEeVM03C6-ROLkeNQo6r-uTO at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> I was researching "orthorectification" a few weeks ago and ran across a
> rather bothersome reference that suggested companies doing aerial
> surveys may keep details of the algorithms for this process proprietary
> in order to protect what they regard as their trade secrets.  I'll get
> the specific reference later, but from my standpoint, hiding details of
> the algorithms means that it is impossible to replicate the details of
> the processes applied to the data or to verify the correctness of the
> algorithms.  In the long term, this means that such data should only be
> published in the Journal of Irreproducible Results.  It also creates
> some rather interesting pleasantries for provenance.
>
> Bruce B.
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> <http://www.lists.esipfed.org/pipermail/esip-preserve/attachments/201011
> 30/c2f2265d/attachment-0001.html<http://www.lists.esipfed.org/pipermail/esip-preserve/attachments/201011%0A30/c2f2265d/attachment-0001.html>
> >
>
> ------------------------------
>
> _______________________________________________
> Esip-preserve mailing list
> Esip-preserve at lists.esipfed.org
> http://www.lists.esipfed.org/mailman/listinfo/esip-preserve
>
>
> End of Esip-preserve Digest, Vol 22, Issue 6
> ********************************************
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.esipfed.org/pipermail/esip-preserve/attachments/20101201/4eb858ca/attachment.html>