MIWP-8 5th virtual meeting 2015-01-22

 

Connection details

 

Agenda

  • Chapters and sections in new release for TG
  • The reference to ISO 19157 quality reports in data specifications
  • Relations to existing draft of TG SDS with TG Metadata
  • Status of work in ongoing tickets (group-leaders)
  • Upcoming meeting dates (Michael Ö)

 Minutes

Suggested chapter numbering

A wikipage is created with suggestion for numbering of new elements
This suggests adding new section to the end of chapter 2 
2.12
2.13
2.14
 
Ine: I'm not sure if we should place the SDS in a separate chapter or if we keep them with 
other elements.
Marc: It would be useful if we keep newly added elements grouped together so that users can find exactly
what they need according to their usecase (updating interoperability element, metadata for SDS etc)
Ine: Should we also separate service metadata ?
Peter: We have different contexts where users are looking for elements and for that reason it would be best to keep new sections as proposed like 2.12, 2.13 , 2.14 etc.
Ini: I agree with Marc but maybe we should split metadata in two chapters, one for datasets and one for services.
Antonio: I would rename chapter 2 "Metadata for discovery", adding a new chapter 3 "Metadata for interoperability". That because they derive from 2 different Regulations
Mig-Sub -Groups: Marc suggests having the metadata grouped by chapters
Marc Leobet (FR): Yes, to be focused on user's needs
James [UK] 2: +1 . need to make it easy as possible for reader - its tricky enough
Michael Ö: Add comments and further suggestions to the ticket for this.
 

Use of ISO 19157 in Data specifications

Michael Ö: This is still an open issue. We are trying to find a meeting with JRC to get the background 
for why they are added into technical guidelines
Marc: I don't see the need for using ISO 191157. We have currently no tools for this.
If we start using this standard we will get a large gap between technical guidelines and current knowledge of users. We should be closer to the word of the regulation.
Peter: There was an update in last versions for all specifications to follow ISO 191157 instead ISO 19115.
We should now focus on elements in ISO19115. There are no elements that are mandatory 
Michael Ö: I have no intention that we should include the ISO 191157 standard. But for the user that wants to include these quality elements into metadata will have problems mixing ISO 19115 and ISO 19157. For this they will have to create two separate metadatarecords.  
Peter: Our catalog-system does not support ISO 191157 and there are no plans to include this in the short future.
James [UK] 2: yes 19157 tooling (and schema) still too immature 

Marc Leobet (FR): +1 to James
Antonio Rotundo (IT): Is it still at draft status or is it published?

Marc Leobet (FR): +1 to Peter : stay pragmatic

James [UK] 2: +1 X 2!

Ine de Visser (NL): +1 to Peter

Martin Seiler, DE: +1 @peter

James [UK] 2: I agree to use existing. Keep it simple without additional extensions. agree QoS is a bit of  asqueeze but need to balance against simplicity
 

Comment from Marc 30.01.15 : I've heard in the end of that discussion " we will keep ISO 19157 out of our scope", am I wrong?

Existing Technical guidelines SDS

These technical guidelines contains mostly metadata.
Discussions on MIG-t suggested moving all content to TG Metadata

 

Metadata for SDS

Ini: There are one major issue here.
In the original TG SDS there is a suggstion of doing an extention to ISO 19115.
There are though possibility to map the requirements to existing elements.
 
Michael: If we create an extention for ISO 19115 that is specific for Inspire our tools will be even more
specific for Inspire.
Peter: There are some elements which does map to existing elements. We should try to add new elements to existing elements and if some few elements does not have an exact match we should anyhow consider to hide these elements in already existing elements in ISO1915 and try to find the best match anyhow.
Ini: The only element that needs extension seems to be Quality of Service that could be hard to match.
James [UK] 2: I agree to use existing. Keep it simple without additional extensions. agree QoS is a bit of  asqueeze but need to balance against simplicity
Michael Ö: I think it could be possible to use existing qualityreports and extend with an additional DQ-Element 
Marc: We need to separate Invoce from Invocable service
Michael Ö: It would be good if larger group in this team could check current tickets in Redmine.
Michael Ö: (We had some technical problems here with clipping sound)
Marc: We have to elaborate on the discussions on Accesspoints and EndpointURLs for SDS what is actually mandatory according to the regulation.
Marc Leobet (FR): It is obvious for me we have to clarify what an endpoint is
Michael Ö: Yes Marc I agree on that we should defined it better
Michael Ö: Also a discussionon the scope of SDS. What services should be documented with SDS Metadata.
Are all services that works on data that are included in Inspire required to be documented, how can we defined thoose services?
James [UK] 2: we had same question about scope needing clarified
Martin Seiler, DE: same question here and we also asked the comission for clarification already. No answers.

 

MIWP-8 (D) Integrate metadata for interoperability into the Metadata TG

Peter: One the six elements are Topological consitency names two ways: descriptive or quantitative.
We have tried to match the descriptive way to ISO 19115 but that will be problematic. We have not found a good way. Besides from this element all other elements are quite clear. 
James [UK] 2: yes. we have a template with some questions to complete
James [UK] 2: we need an example for Temporal ref.
James [UK] 2: Issue relating to Class for Spatial representation and more discussion I think on Topo Consistency
James [UK] 2: 19115 raises specification and conformity  issues
Marc: We done some work on temporal reference and topological consitency in French Guidelines
Peter: Marc and James can you add your suggestions for temporal reference into the wiki?
James: It would be great if a wider group checks the document.
 
MIWP-8 (I) Language neutral identifiers
Michael Ö : I should have setup a codelist in registry on this to show how the values could managed 
No work is yet done here. Some comments are added to wiki.
 

 

MIWP-8 (L) Unique Resource Identifier

Martin: Wiki-page is modified with two use-cases. The element is changed now to MD_Identifier.
I can start to add this into the actual TG document.
Michael Ö: Are there work going on related to handling identifiers related the needs for Open data and Inspire together. Or do we need to handle two separate identifiers?
Martin: We keep the identifiers we have in the SDI and have separate identifiers for Open data portals.
Christine Gassner (AT): in Austria inspire and open data are handled very close. profil.AT try to have a synoptic point of view as well on identifiers
Marc: In France its easy : the French portals don't use unique identifiers.
Christian: There is a report by JRC on the mapping between Open data to Inspire 
 
 

MIWP-8_(M)_Coupled_resources

Michael: We had dicussions on how its used today.
Michael: Two usecases are added to wiki where the coupled resources are used.
From the use cases point it is the uuidref that has to be mandatory.
xlink:href could be optional. But it is the mainly used implementation in member states.
Paul: Related to monitoring we need to solve the use-case to find if datasets are accesible through a service.
Martin: Can you Paul write down the use case for the Monitoring needs of Operates On.
We need to check the TG Guidelines for discovery and see if its clearly specified how Operates On should be managed in a CSW. 
Michael: In Geonetwork the default mapping of OperatesOn is against the uuidref-attribute
Martin: But that can be different between different implementations. We need to check what technical guidelines for discovery are requireing
Kristian: I have created a document to redmine describing my thoughts and ideas on Coupled resources.
One point i rasied is the use xlink:href which against the ISO-model. If we use xlink:href we should not use 
GetRecordById. One issue related to using only uuidref is to know what catalogue to find the record that describes
the resource with the uuid set in uuidref.
Martin: It would be good to make a recommendation on best practices. The probability that existing implementations wil be changed is low.
 

MIWP-8 (E-1) Integrate Theme specific metadata

Peter: This is same approach as for the six interoperability elements.
We have some issues marked with yellow in the excel-sheet on elements that are not fully clear. 
These elements should be checked by a larger group.
We have same problems here with the elements in ISO 191157, that does not fit with current TG.
Ine: I will update table next week

 

MIWP-8 (A) Conditions applying to access and use

Marc: The Michael's proposal is very interesting as it allows national consensus for interpretation of ISO19115, avoids a long debate fo fix an European interpretation and lets opened the possibility to post-process an harmonisation. We check the possibility to add virtual CSW front to process metadata before it enters Inspire portal.
Peter: I wonder how you will handle a CSW-transform that can understand what element to transfrom for useLimitation.
Michael: Lets discuss this on the wiki.
 
Comment from Marc 30.01.15: most of the time, data producers do not know how to fill clearly the ISO19115 fields. The result is to hardly recognize opened datasets. We have the legal framework to force State's data with unclear conditions toward open data licence, without to have to come back to the data producers. 
 
 
 

Actions

Added Description Assigned to Status Done
2015-01-22 Read through the suggested mapping of SDS elements. And return with feedback All    
2015-01-22 Check Kristians document on coupled Resources All    
2015-01-22 Check with JRC on status forISO 19157 implemementions Michael Ö    
2015-01-22 Each country should report (by email to Michael Ö) on how they have implemented OperatesOn in Service metadata:
are you using xlink:href ?
uuidref ?
Are there a queryable in their CSW for OperatesOn and ResourceIdentifier? What are they mapped to ?
All    
2015-01-22 Use case for OperatesOn in monitoring. Describe further how the Monitoring group would like to query the CSW. Paul