Meeting details:

• Date: 25 July2014
• Venue: Video/teleconference
• Meeting started: 10:00 AM
• Meeting ended: 12:00 PM
• Meeting Chair: Nigel Ward

Attendees:

1. Anne Stevenson (CSIRO)
2. Nigel Ward (QUT)
3. Dave Connell (AAD)
4. Daniela Nastasie (UniSA)
5. Irina (GA)
6. Jingbo Wang (NCI)
7. Marianne Browne 
8. Peter Walsh (UTas)
9. Gerry Ryder (ANDS)
10. Julie McCulloch (ANDS)
11. Liz Woods (ANDS)
12. Cel Pilapil (ANDS)

Guest:

Melanie Barlow (ANDS)

Apologies:

Conal Tuohy (VeRSI), Gavan McCarthy (UNIMELB), Neil Godfrey (CDU), Steve Androulakis (Monash), Simon Porter (UNIMELB), Adrian Burton (ANDS)

Agenda:

1. Welcome - Nigel Ward

2. Short introduction –all RAB members

3. Discussion of RIF-CS change proposal and briefing papers (ANDS, RAB members)

4. Other business - Nigel Ward

5. Date and time of next meeting

D I S C U S S I O N :

A. Short introduction - all RAB members

B. Role of the RAB - briefly discussed by Nigel Ward

C. RIF-CS Schema recommended changes:

1.     Services over data (Discussed by Melanie Adamson)

This proposal aims to:

• Allow RIF-CS users to encode information about
• whether data is openly internet accessible
• the available URL’s for accessing data
• the available services to access the data
  • From a schema perspective this proposal proposes:
    • Addition of optional /@target and /@visible attributes for the electronic element (within  location/address element) to indicate  a direct download URL or a landing page.
    • Addition of optional child elements title, notes, mediaType and byteSize  to provide more information about the /@target (for directDownload)
    • Removal of ‘wsdl’ from electronic type vocabulary

Discussion:

• Melanie first presented the RAB with a mock-up in Research Data Australia, showing of how RDA collection record view would look like with the implementation of the proposed changes
• Daniela expressed her support for this proposal. She mentioned that UniSA is currently undertaking a project towards 'open access'
• Melanie went ahead and presented the proposal
• Nigel went around the table to ask each RAB member present to provide their comment on the proposal
• Anne asked about how the rights element would support the proposal. She wanted to know how this will work if users search for open access datasets, how is it going to be done via 'directAccess' through electronic address element.
• Melanie then clarified that this is just an initial implementation around open access.  How the rights element will be used to present 'open access' will be analysed separately.  If someone wants data to be downloadable directly but didn't want it visible in RDA then the /@visible attribute will be used.  If accessRights element has clear and direct information about how the data can be accessed then, this attribute will not be necessary. 
• On the other hand, Peter raised his concern about possible confusion it would raise if researchers search for open access and other providers may have already implemented or adapted the change.  ANDS should be very careful as to how the best practice or the advice around the usage of the rights or accessRights element will be used.  
• Melanie clarified that this is just the first phase of the implementation around 'open access'.  The idea is to show per collection record view when a record can be directly downloaded via the /@target attribute or manipulated via a service using a combination of the collection-service relationship 'isSupportedBy' and the value in the url sub-element.
• Peter then raised a possibility that there may be records within RDA that use 'isSupportedBy' in a way that does not support 'open access'.
• Nigel then presented that idea that they may be 2 scenarios here :
  • online access - that can be addressed by the proposed change to the location/address/electronic element (/@target attribute)
  • open access - something that needs further analysis
• Jingbo then shared with the group that for them, one of their concerns is the ability to control access to the data and who accessed it.  She then asked if it was possible to present a license and the user was asked to accept the license prior to accessing the data.
• Melanie and Nigel clarified that, Research Data Australia, being a repository of research data descriptions only, does not store the data nor control access to it.  Ownership and management of data is still with the provider.  So, in essence, if the license if presented to the user via a service or via the link to the data, then, access restrictions will be invoked as per provider’s specifications.
• When it was Irina's turn to comment on the proposal, she expressed her support and said that they very much welcome the development of a functionality that supports 'open access'.   She is, however, interested in statistics - a report showing who has downloaded or accessed the data via Research Data Australia. Also, she asked for clarification where the metadata about the service would be if the link to access or download data is provided via a service. 
• Melanie explained that if a registry object for a service exists within Research Data Australia, then the title of the service links to the registry object itself.  If a registry object does not exist, a page displaying the relationship between the service and the collections/datasets will be available in RDA.
• Dave asked for clarification about the options in the issue paper as to whether the RAB had to choose an option for implementation.
• Cel clarified that this one is different and RAB would not be ask to choose but rather, items 2 & 3 are the proposals which are being presented for approval. Item 1 would need a separate analysis and recommendation. 
• Daniela pointed out that the community should have a common understanding of to what 'open access' means. For her, it means data is freely available.  For researchers, data sets should be described properly and the accessRights should have granular or detailed information about how the data can be accessed.  Daniela and her team are willing to get involved in this initiative by ANDS and that they currently working on making their datasets openly available.
• Gerry, on behalf of Adrian, expressed her support for the proposal. Being involved in documenting the Content Providers Guide and the Best Practice Guidelines, her main concern is around definitions - what open access means, how it is applied; and be able to provide clear guidance on encoding.
• Liz, on technical point of view, mentioned that if it was only faceting and clear display in Research Data Australia, then the proposal using the @target attribute and the additional optional information, together with the service-collection relationship, will be enough to achieve this purpose.  We, however, need to interrogate the records in Research Data Australia to get more information as to how many of these collection-party relationship use the relation type 'isSupportedBy' and understand how they are used by our providers.
• Hearing everyone's feedback, Nigel then proposed that:
  • This RAB meeting would only be for items 2 & 3 - online access to data
  • ANDS to get back to RAB about 'open access' proposal, even if it means an addition of 'rights element vocabulary only and/or enhanced documentation (clearer definition and guidance).
• Going around the table, everyone endorsed Items 2 & 3.

Decision:  RAB endorsed this proposal.

 Actions:

• Re-write proposal to only include items 2 & 3
• ANDS to write a new paper around 'open access'’
• Get statistics or data re: usage of ‘isSupportedBy’ relation type  (collection-service relation) in Research Data Australia

2.     Services over data (Discussed by Melanie Adamson)

This proposal aims to:

• Provide an option of providing a PubMedID identifier when describing publications using the relatedInfo element.  
• From a schema perspective this proposal proposes:
  • Add ‘PubMedID’ in the relation type vocabulary, schema structure not changed.

Discussion:

• Gerry presented the paper re: addition of PubMedID as new identifier type for relatedInfo element:
  • It was driven by ANDS work for Thomson Reuters Data Citation Index
  • PubMedID is an optional type to support related publication

Decision:  RAB endorsed this proposal.

D.    Other Business:  None.

E.     Date and time of next meeting:

• Depending on how fast ANDS could document the proposal around 'open access', a new RAB meeting, if necessary will be scheduled by ANDS.
• Cel discussed that normally, RIF-CS changes happen sometime around September/October so that the schema is ready by November, the major release where the UI (RDA) will be ready to support the new schema
• If a meeting is necessary, one will be scheduled by August.

Action:

• ANDS to schedule the next meeting, if necessary

ACTIONS: