Hi Stu and list,
The success of the list testifies to the need, that's for sure! And I am extremely supportive of the directions Stu is talking about, I hope others will jump in with their support or suggestions.
With regard to the suggested ideas below, I find them exciting and terrifying at the same time. Exciting because of all the great information that is being exchanged, and the great opportunity the list represents. Terrifying because ... well, MMI [1] is already trying to collect great information like this, and it's so easy to get behind!
I noticed the "Current element sets" list included MMI. "MMI" is the Marine Metadata Interoperability project, and we don't generally try to create standards or specifications. We try to capture everything about science metadata (focusing on things applicable to marine science) that is well known. And, we try to provide useful metadata guidance, most of which is science domain neutral.
So far we have about 2500 to 4000 public pages. Of these maybe 500 or 1000 are references like the ones below, another 200 to 500 are guidance documents that we wrote, directly addressing topics like your Linked data bullets. Another 1000 or so may be miscellaneous pieces of information (event listings, job opportunities, workshop documents, etc.).
The point is, we've been working on this site for 5 years now, spending close to $1,000,000 making these resources visible, and lots more effort has been volunteered. Anyone can contribute content or join our teams and projects, and we have a thriving community of site contributors, many of whom then collaborated on other projects.
So we are familiar with steps 1, 2, and 3 below. And while I hated to admit the dollar amount (why haven't we done even more?!), I hope it gives everyone an idea of the scope and effort required to make this kind of information come together effectively. And yet there is plenty left to do.
So, on the one hand, much of our work seems directly useful to the DC Science and Metadata Community (and community). But on the other hand, the DC science community has a wider scope, audience, and level of possible participation than the ocean science community alone can generate. Having the DC science community looking to move things forward is really exciting.
I would love to see the energy on this list leverage the effort that's gone into MMI, to make the most/best possible use of both. I can imagine all sorts of ways to do that, but it will be most interesting to see what people on this list see as needs, and want to take on. To whatever extent that takes advantage of MMI's work to create better solutions for all science communities, we're on board.
Thanks for your time, I apologize for the long post but hopefully you can appreciate my enthusiasm.
John
P.S. Anyone is welcome to email me privately with comments on MMI, whether good or bad, if they aren't relevant to the discussions. But I won't mind if you critique MMI on the list either -- it's all helpful.
On Feb 26, 2009, at 10:27 AM, Stu Weibel wrote:
Hi, Folks,
It is gratifying to see that we’ve got well over 100 people signed up to DC-Science, and there has been some interesting discussion right off the starting line.
It is easy to imagine all this discussion could be fragmented and lost in the archives of the list. What might we do with all this discussion to extract useful content that helps us solve problems? My strawman proposal:
- Establish a community WIKI that carves up the topics into digestible chunks.
- Assign an editor for each of the chunks, whose responsibility would be to track the conversations on the list and extract important topics, links, use cases, and actions to take
- Shape the wiki into a useful community resource
The chunks, based on our early discussions, might be something like:
Linked data – the state of URI identification of controlled vocabularies
- Useful conventions should apply to URIs
- the benefits and problems with semantics in identifiers
- The state of enabling technology for semantic web applications
- Circumstances under which string data is acceptable or desirable as metadata values (without URIs)
Current element sets maintained by various different organizations covering various types of data sets, and their relationship to DC metadata
- ISO 19115
- NASA DIF
- MMI
- FGDC
- EML
- GML
- Darwin Core
- DC-Collection-level data
Data models for Scientific Data Sets
- Schemas for data
- Terminology and granularity of data- sets, collections, series, etc.... The
- What IS a dataset?
- What is an appropriate level of granularity for identification? For metadata creation?
- Data Formats and data types
Functional Categories of metadata:
- discovery
- advanced searching
- interpretation for re-use
- preservation
- semi-automated data integration
These may not be precisely the right categories, and certainly other topics will emerge.
What I am concerned about is that we use the list and organize resources in such a manner as to actually help people solve their problems. By all means amplify this list with your own ideas