Discussing "Contextual Metadata"

Where we discuss anything related to the RDA (catchall)

Moderators: Leif.Laaksonen, SaraPittonetGaiarin

Discussing "Contextual Metadata"

Postby brigitte » Sun Feb 24, 2013 11:54 pm

Attached is the Case Statement for "Contextual Metadata". A "Contextual Metadata" slot is reserved at the BoF session in the Gothenburg Plenary to continue discussion f2f, where here we hopefully manage to get the discussion going to prepare this BoF session: http://forum.rd-alliance.org/viewtopic.php?f=5&t=60

A first discussion point may be the proposed use-cases, which are central to the "Contextual Metadata" approach:

Proposed Use Cases:
(1) Output-Reporting to Funders
(2) Exchange of Information on Research Activity between Universities
(3) Management of the Research portfolio of a University by a Research Manager
(4) Discovery and Re-use of Datasets for other Purposes

Are these suggested use cases appropriate or of relevance? Which priority with respect to feasability would one give them taking into account available material and experience of group members? Do we need to define them further? What are your views? Where do we need further clarification?
You do not have the required permissions to view the files attached to this post.
brigitte
 
Posts: 12
Joined: Mon Feb 11, 2013 3:44 pm

Re: Discussing "Contextual Metadata"

Postby msicilia » Mon Feb 25, 2013 9:23 pm

Dear Brigitte,

Many thanks for the document to start the discussion. In my opinion, the use cases are relevant and really set the context of research as a set of activities that require consideration from different viewpoints.

Just a small suggestion, one of the cases is "Discovery and Re-use of Datasets for other Purposes". Here I think we could specify a bit further the "other purposes" to include some additional relevant viewpoints. As the other use cases are more institutional or manager-oriented, maybe for this one we could include a more researcher-centric perspective. For example: "discovery and reuse of datasets for experiment replication" (this is only a suggestion, there are other scientist-focused possibilities). As reproducibility is an important element of the scientific method, this discovery of datasets would enable better possibilities for scientists to locate "similar experiments" for the purpose of contrastinv with their own or to do meta-analysis eventually. This is in coherence also with a more "data-intensive science" approach for which big data and massive dataset sharing and reuse provides unique new opportunities. This is also in connection with current projects that aim at creating repositories of scientific workflows or to exploit shared semantics and linked data as advanced means for discovery and reuse of datasets in a global context.

Best regards,
Miguel-Angel Sicilia
msicilia
 
Posts: 1
Joined: Mon Feb 25, 2013 4:32 pm

Re: Discussing "Contextual Metadata"

Postby parsonsm » Mon Mar 04, 2013 4:26 pm

The Contextual Metadata group may be interested in work being done by a broad coalition of Earth Science data organizations in the US looking at developing a "provenance and context content standard" http://wiki.esipfed.org/index.php/Prove ... t_Standard

Their focus is on Earth science data especially remote sensing data, but I suspect there is some overlap with this group. For example, they have made good use of the developing W3C Prov spec.

cheers,

-m.
Mark A. Parsons
RDA/US
User avatar
parsonsm
 
Posts: 18
Joined: Sun Nov 04, 2012 8:44 pm

Re: Discussing "Contextual Metadata"

Postby brigitte » Thu Mar 07, 2013 12:50 pm

Dear Miguel-Angel,

thank you very much for your add-on and your feedback - especially that the use cases are relevant. Your point concerning the last one is a valid one - and the more appreciated - especially as I know you have a lot of experience with processing data and teaching/supervising students to do so. I fully agree that the first three are more institutional / managerial and a lot of material is already available for collection and aggregation from where we can start. The proposed use cases also have an implicit order with respect to priority aiming at feasibility and delivery from what has already been done (and surely this is a point for discussion in the BOF).

The use case (4) to which you point could indeed be more researcher-focussed - a very good point - there may be further purposes for later as well (there may always be) - but taking the researcher's perspective is certainly a relevant one - and as you say highly relevant with "data-intensive science" and current projects aiming at an infrastructure for advanced discovery and global sharing. We have the following cases now.

Proposed Use Cases:
(1) Output-Reporting to Funders
(2) Exchange of Information on Research Activity between Universities
(3) Management of the Research portfolio of a University by a Research Manager
(4) Discovery and Re-use of Datasets for Experiment Replication

Thank you very much!
Brigitte
brigitte
 
Posts: 12
Joined: Mon Feb 11, 2013 3:44 pm

Re: Discussing "Contextual Metadata"

Postby brigitte » Thu Mar 07, 2013 1:00 pm

Dear Mark,

thank you very much for your interest and contribution! I will certainly take it into account with preparations for the BOF session - which are being prepared - and the Prov Spec is in the list already.

Will you attend the plenary?

Brigitte
brigitte
 
Posts: 12
Joined: Mon Feb 11, 2013 3:44 pm

Re: Discussing "Contextual Metadata"

Postby brigitte » Thu Mar 07, 2013 1:03 pm

Putting forward the feedback by Keith Jeffery for the BOF session (see here: http://forum.rd-alliance.org/viewtopic.php?f=5&t=60)

Re: BoF session on "Contextual Metadata"
by KeithJ » Mon Mar 04, 2013 9:19 am

All -
I believe we have some interesting challenges in this BoF and hopefully approved group:
1. how we interact with the other groups and particularly the proposed metadata group (which seems to me to be a very wide topic);
2. what architectural model do we adopt - for example the 4-level model proposed by various organisations/projects of: user metadata (user profile, preferences, authorities, responsibilities), data metadata (describing information sources), process metadata (describing software components and their properties) and resource (computers, networks, storage, detectors) metadata? - as used in GRIDs, CLOUDs and other e-infrastructures;
3. how do we relate discovery metadata, and detailed (schema level) metadata with contextual metadata in all of the above models;
4. what is the relationship of contextual metadata to restrictive metadata (rights, security, privacy, costs/charges etc);
5. what is the relationship of contextual metadata to navigational metadata (how to reach the object(s) of interest from the contextual metadata and associated restrictive metadata (expressed as constraints on the navigational metadata) etc?;
6. what is the relationship of contextual metadata to preservation metadata (e.g. using the OAIS framework)?;

There is plenty to discuss!
Keith


My response:

Re: BoF session on "Contextual Metadata"
by brigitte » Thu Mar 07, 2013 6:37 am

Hi Keith,

good to know you will attend. We have already been in contact with the Metadata group (Jane Greenberg and Rebecca Koskela) and there is certainly room and willingness for collaboration and joining forces - on both sides - and some people in both groups seem to know each other - that is even better. The Gothenburg meeting will certainly be a good opportunity for a face-to-face chat and further discussion in this respect. And it is not long to go.

Thank you for your contribution, which I will take forward to the discussion area, so we have everything at one source, i.e. that makes it easier to coordinate discussions. Please see here: viewtopic.php?f=2&t=67

For sure there is a lot to discuss. The BOF session is being prepared and you will find more information about it shortly from within the just mentioned discussion area.

Thank you for your contribution!



The BOF session is being prepared and I will post more on that.
brigitte
 
Posts: 12
Joined: Mon Feb 11, 2013 3:44 pm

Re: Discussing "Contextual Metadata"

Postby paolo.manghi » Tue Mar 12, 2013 10:18 am

msicilia wrote:Dear Brigitte,

Many thanks for the document to start the discussion. In my opinion, the use cases are relevant and really set the context of research as a set of activities that require consideration from different viewpoints.

Just a small suggestion, one of the cases is "Discovery and Re-use of Datasets for other Purposes". Here I think we could specify a bit further the "other purposes" to include some additional relevant viewpoints. As the other use cases are more institutional or manager-oriented, maybe for this one we could include a more researcher-centric perspective. For example: "discovery and reuse of datasets for experiment replication" (this is only a suggestion, there are other scientist-focused possibilities). As reproducibility is an important element of the scientific method, this discovery of datasets would enable better possibilities for scientists to locate "similar experiments" for the purpose of contrastinv with their own or to do meta-analysis eventually. This is in coherence also with a more "data-intensive science" approach for which big data and massive dataset sharing and reuse provides unique new opportunities. This is also in connection with current projects that aim at creating repositories of scientific workflows or to exploit shared semantics and linked data as advanced means for discovery and reuse of datasets in a global context.

Best regards,
Miguel-Angel Sicilia


Agree with the observation. We may also refer to the 6 R's here: http://wiki.myexperiment.org/index.php/Research_Objects and consider the idea of linking datasets and devices/facilities in the CERIF model. To some extent CERIF could be mapped onto myexperiment workflows. Too ambitious? Out of scope?

Paolo
paolo.manghi
 
Posts: 2
Joined: Thu Oct 25, 2012 3:21 pm

Re: Discussing "Contextual Metadata"

Postby brigitte » Tue Mar 12, 2013 10:51 pm

Hi Paolo, all,

thank you very much for your thoughts - good you like the comments made by Miguel-Angel.

Just to recall for everybody. So far we have the following use-cases:
(1) Output-Reporting to Funders
(2) Exchange of Information on Research Activity between Universities
(3) Management of the Research portfolio of a University by a Research Manager
(4) Discovery and Re-use of Datasets for Experiment Replication


Agree with the observation. We may also refer to the 6 R's here: http://wiki.myexperiment.org/index.php/Research_Objects and consider the idea of linking datasets and devices/facilities in the CERIF model. To some extent CERIF could be mapped onto myexperiment workflows. Too ambitious? Out of scope?

Paolo


Exciting definitely - depends on how much time / resources we need to invest to define or identify the basic objects behind the use cases - that I think needs to be clear first. If there is enough re-usable material so we can proceed to workflows - would be great (that is my view of course). We have to see about resources and timing - could be a next step - what do others think?

I take the freedom to mention here some items that have been discussed via email. A suggestion by Jian Qin to take into account quality and verification with datasets and a response by Angus Whyte - that the verification aspect may be reflected already with use-case (4), but may need adjustment in its wording.

Quality will thus be a cross-use-case discussion item for our BoF session in Gothenburg. Let's see if we find additional ones before the Plenary :)

Thank you all for your active engagement!

Brigitte
brigitte
 
Posts: 12
Joined: Mon Feb 11, 2013 3:44 pm

Re: Discussing "Contextual Metadata"

Postby brigitte » Tue Mar 12, 2013 11:30 pm

Another aspect that will certainly be important to discuss within the BoF session in Gothenburg, is: how we connect with the "Metadata" group proposed by Jane Greenberg and Rebecca Koskela (http://forum.rd-alliance.org/viewtopic.php?f=2&t=22).

In addition to the proposed use cases:
(1) Output-Reporting to Funders
(2) Exchange of Information on Research Activity between Universities
(3) Management of the Research portfolio of a University by a Research Manager
(4) Discovery and Re-use of Datasets for other Purposes

we therefore currently have the following additional aspects for discussion (see also previous posts):
- quality
- verification
- research objects
- connecting with other groups
- architectural model
- relationship discovery metadata with contextual metadata
- relationship detailed schema metadata with contextual metadata
- relationship restrictive metadata with contextual metadata
- relationship process metadata with contextual metadata
- relationship navigational metadata with contextual metadata
- relationship preservation metadata with contextual metadata

Any further wishes / comments - please add on!
brigitte
 
Posts: 12
Joined: Mon Feb 11, 2013 3:44 pm

Re: Discussing "Contextual Metadata"

Postby joncr » Wed Mar 13, 2013 12:16 am

From the library perspective, I would just add two comments: first, that it's very hard to convince researchers to add enough metadata when they deposit data in a repository to make the 6Rs possible, so new incentives will be necessary (and may be possible as habits and expectations change). Second, in seeking to foster reuse of data it will be key to push data sharing much earlier in the research process. Libraries and repositories have too often been seen as the final resting places for archival content, while replication and reuse will in most cases need to happen while data are fresh and software or research methodologies have not changed so much that replication would be difficult if not impossible. -- Jon
joncr
 
Posts: 1
Joined: Tue Mar 12, 2013 1:33 pm

Next

Return to RDA Discussion Area

Who is online

Users browsing this forum: No registered users and 1 guest

cron