[Esip-dds] A data life cycle model

[Ruth Duerr]

Well… these last comments have opened up the can of worms I was hoping could stay closed.  The facts of the matter that there are strengths and weaknesses to every data model and I don't think you could ever get more than say 2 or 3 people to agree on a single one without a lot of effort.  Since our goal for picking a data model was to just provide something to measure existing work against, all the one we choose needs to be in my opinion is just good enough that all of the various issues can be tagged to it somewhere.  That said, I agree with everything said so far, and more than that, in order to paper over the archive vs current access issue, I prefer to talk about enduring access, as the things you need to do to ensure that something is useful today are all needed to ensure that it is still useful tomorrow or a few/many years from not (and a yes you need more to make sure it is still going to be useful a bit down the road mostly because the user communities change with time and understanding that and accommodating it is an ongoing process).  

Talk to you all shortly.

Ruth

On Feb 28, 2013, at 12:06 PM, Joe Hourcle <oneiros at grace.nascom.nasa.gov> wrote:

> 
> 
> On Thu, 28 Feb 2013, Ramdeen, Sarah wrote:
> 
>> Hello Everyone,
>> 
>> I am a bit behind in my emails but I wanted to add some comments about the lifecycle model.  Echoing Erin's comment, I have always preferred the DCC mode (http://www.dcc.ac.uk/resources/curation-lifecycle-model) because of the different loops represented in the diagram, but I think that is merely because it is an easy to understand visual and not because it is a better model!  That said, there are a few things I think the proposed model is missing.  These are just suggestions of course - things I thought about after a first pass of looking at the current model.
> 
> I'm just getting caught up as well, as I only joined the list the other day.
> 
> It's going to take me a while to remember all of the various models that I've seen over the years, but I agree -- I think the straight line of processes isn't as accurate of a representations as some of the ones that suggest that it's an iterative process. (media refreshes, repackaging as standards change, evaluate if it should be discarded*)
> 
> There was also a proposed extension to the DCC model to insert a couple of more items ('User Experience' and 'Knowledge Enhancement'):
> 
> 	http://www.ijdc.net/index.php/ijdc/article/view/69
> 
> I don't know that they're necessary, as I view 'knowledge enhancement' as part of 'Curation', and UX as something off to
> 
> 
> 
> 
>> 1) In the data life cycle model section there is not a step for appraisal or collection development.  This is implied with collect I guess but there are usually some decisions made about what should be part of a collection and not everything collected is kept.  This is different than deciding that something is no longer of value - but that assessment once something is donated or initially collected to determine if it belongs as a permanent part of the collection.
> 
> I'd say that the process is a little different than your standard collections development -- right now, we're prepping for a senior review -- if we can't show that the material's still of value to the community, then it gets transfered to an archive with a lower cost of storage.  I could see that in some cases, an archive might maintain different classes of storage, and move stuff from online to nearline or offline / offsite.
> 
> When we're dealing with ingest, it might be that we determine that something that was given to us isn't actually appropriate for the community that we serve, and we might try to transfer it to some other group.
> 
> And both of these happen at levels above me right now.  (although, I've been warned that next year, we have to look into our archive dealing with long-term preservation).
> 
> 
> 
>> 2) In this same section it does not implicitly discuss access and use either.  Again, probably implied with discover, but I think that being able to discover is slightly different than being able to gain access/permissions or properly use the materials as well.
> 
> Also reminded me -- back in 2007, there was a workshop headed up by the OAIS authors called 'Science Archives in the 21st Century'
> 
> 	http://nssdc.gsfc.nasa.gov/nost/conf/archive21st/
> 
> There was a pretty big rift between the folks who looked at archiving as being something for the future, vs. those who were more concerned about serving people today.  (I don't want to put words into people's mouths, so I won't say who I think said it, but there was a comment at the workshop
> that if we couldn't make sense of the data today, what good was it to store it for someone else to figure out in the future)
> 
> ... um ... I'm going to stop commenting, as the telecon's in an hour ...
> 
> ... but I agree with Sarah on the other two.
> 
> -Joe
> _______________________________________________
> Esip-dds mailing list
> Esip-dds at lists.esipfed.org
> http://www.lists.esipfed.org/mailman/listinfo/esip-dds