Hi Ruth,

I've tried to answer your questions, sounds like you are having a trying time!

1. Is something like this mapping useful to the community or is it  
silly to even think about it? [I would be willing to post it on the  
PIG wiki; but, reserve the right to publish it in a paper I am  
writing].  If it is useful, would the community be interested in  
reviewing/helping me solidify the draft?


I think that this would definitely be a useful resource for the community to have available. I know that Geospatial metadata is an area in which a lot of people are keenly interested.

2. What is your take on the FGDC vs PREMIS metadata issue?


From my (extremely) quick perusal of the FGDC i would say that it is mainly used/useful for descriptive and discovery metadata purposes whereas PREMIS is all about preservation. They seem like they would compliment each other rather than overlap. In the data archive in which I am working we are using the DDI (Data Documentation Initiative)standard for Descriptive/Discovery Metadata and an adaptation of PREMIS for Preservation metadata in much the same way as I would envisage you using the FGDC and PREMIS.


3. What is your opinion on the question of whether data centres and  
digital libraries, etc. should/could use the same set of standards?


I personally believe that it is virtually self-evident that all data centres/digital libraries etc should Ideally use the same set of standards,simply for reasons of interoperability and/or knowledge sharing.

However my experience is that no matter how comprehensive a metadata standard is, it usually has to be altered to be useful for any particular implementation. For example we have altered both the DDI and the PREMIS standards for our particular usage. As archiving and even metadata standards are still in a relatively nascent stage at the moment it is likely that the metadata standards will become more and more useful and widely implementable as they get improved upon. This has already happened with the DDI, version 2 (which we use)was recently surpassed by version 3 which seems to address a few of the implementation issues.  
The PREMIS xml schemas have tried to tackle this issue by having areas in them in which you can put in whatever metadata elements you which (i.e. you can make them up). This both improves its interoperability (as more people will be able to use the unaltered schema) and detracts from its interoperability (as other users may not understand you particular implementation of it).

Also, the PREMIS standard is more of a reference standard than a strict set of terms to be used. It is best used as a checklist to ensure that you are collecting all of the metadata you need for preservation purposes. So even if you decide not to use the PREMIS schemas, PREMIS could still be of use for this purpose.

If you need to persuade your colleagues of the need for digital preservation generally and preservation metadata in particular i can recommend the On-line Cornell Digital Preservation Tutorial:
http://www.library.cornell.edu/iris/tutorial/dpm/

And this article in popular mechanics has a nice, easily accessible, overview of some of the issues involved in digital preservation:
http://www.popularmechanics.com/technology/industry/4201645.html

I hope this is helpful, if you have any questions or feedback please feel free to get in contact.

Kind regards,



Euan Cochrane

Statistical Analyst,
OSRDAC (Official Statistics Research and Data Archive Centre)
Statistics New Zealand
[log in to unmask]



Ruth Duerr <[log in to unmask]>
Sent by: PREMIS Implementors Group Forum <[log in to unmask]>

07/07/2007 11:51 AM
Please respond to
PREMIS Implementors Group Forum <[log in to unmask]>

To
[log in to unmask]
cc
Subject
[PIG] PREMIS and FGDC metadata question





I'd like your opinions on a few issues.  I work at a data center that  
primarily archives and distributes remote sensing data about the  
environment and as such I am probably one of, if not the only,  
science data manager that is interested in PREMIS metadata.  My  
reasons for being interested are:

1. I am a fan of the OAIS reference model
2. We need to use something to manage information about our archive
3. I hate reinventing the wheel (though am willing to balance wheels  
and add new spokes as need be)
4. It should eventually help interoperability allowing archives to  
easily back each other up through the exchange of metadata and data
5. It should eventually help harmonize the directions digital  
libraries and data centers are headed (I'd love to add the GIS  
community to this too), thereby providing a more seamless transition  
from data to information and knowledge, if not exactly to wisdom...

I've noticed that I get a lot of push back on using PREMIS, not just  
internally; but, also from my fellow science data managers  
elsewhere.  Some of that may simply be resistance to change or the  
infamous "not invented here" syndrome. It also may be partly that  
there is a lot of pressure to head in the direction of less metadata  
rather than more (i.e., I hear sentiments like "metadata isn't the  
solution - we need a better hammer" a lot).  One question I'm often  
asked is why anything beyond FGDC metadata is needed (almost all of  
my and my colleagues' data is documented at the data set level by  
FGDC metadata).  My answers about storage information, fixity,  
preservation events and agents, and rights are routinely met with  
statements like "FGDC can do that."  That has always seemed strange  
to me since the FGDC standard was specifically developed to contain  
metadata to support the following:

"The information included in the standard was selected based on four  
roles that metadata play:
- availability -- data needed to determine the sets of data that  
exist for a geographic location.
- fitness for use -- data needed to determine if a set of data meets  
a specific need.
- access -- data needed to acquire an identified set of data.
- transfer -- data needed to process and use a set of data." - from  
CSDGM, 1999

Not one of these purposes is to ensure the long-term preservation of  
data.  As such, my first thought has always been that FGDC and PREMIS  
metadata should be orthogonal - in other words, there shouldn't be a  
lot of overlap between the standards.  It that is the case, then it  
seems to me that it would make a lot of sense to use both standards  
simultaneously - FGDC to deal with external user access, PREMIS to  
deal with preservation needs.  Since I've gotten so much push back on  
this, I decided to see how much overlap between the two standards  
there really is.  I've attached a very rough draft of a PREMIS to  
FGDC mapping and am contemplating drafting the inverse FGDC to PREMIS  
mapping in addition.  The map was drafted for our internal wiki - so  
lot's of the comments are NSIDC specific.  My own impression is that  
there is a bit more overlap than I was expecting; but, that I had to  
pound those square pegs pretty hard to get them to fit in those round  
holes.

My questions for the group are:

1. Is something like this mapping useful to the community or is it  
silly to even think about it? [I would be willing to post it on the  
PIG wiki; but, reserve the right to publish it in a paper I am  
writing].  If it is useful, would the community be interested in  
reviewing/helping me solidify the draft?
2. What is your take on the FGDC vs PREMIS metadata issue?
3. What is your opinion on the question of whether data centers and  
digital libraries, etc. should/could use the same set of standards?

Thanks a bunch (in advance),

Ruth Duerr
NSIDC Data Stewardship Program Manager and MODIS/PARCA Data Coordinator


 

Visit Statistics New Zealand at the Small Business Expo,

TSB Bank Arena, Queen's Wharf, Wellington, 18-20 July 2007.

 

====================== Correspondents: Please Note =====================

The information in this email, and any files transmitted with it, is

confidential and is for the intended recipient only. If you receive

this message in error, please phone us toll free on 0508 525 525, or

notify us via [log in to unmask]

 

The content of any email entering or leaving Statistics New Zealand is

automatically scanned, and may be opened and read by security staff.

 

Statistics New Zealand makes reasonable efforts to ensure that its email

has been scanned and is free of viruses. However, Statistics New Zealand

can make no warranty that this email or any attachments to it are free

from viruses.

========================================================================