Meetings/20070606/JISC-DigitalDeluge-day2
From ImageWeb
Contents |
JISC Digital Deluge repositories conference - day 2
- Miscellaneous - Miscellaneous notes and meetings page in wiki
- Meetings/20070605/JISC-DigitalDeluge-day1 - day 1
- Meetings/20070606/JISC-DigitalDeluge-day2 - day 2
- http://www.jisc.ac.uk/events/2007/06/repositories_conference.aspx - conference page
Drummond Bone - President UUK, VC Liverpool
Nothing fundamentally new about repositories. (This guy has English Lit background.)
UK produces: 9% sci pubs, 12% citations, 1% population (?) (hmmm, what if BRICS factored out?)
Why are repositories a Good Thing?
Managerial
Bundling research output increases visibility/prestige. "Brand presence" stuff. (This all seems very corporatist, no recognition of actual researchers here)
Exploit marketing potential of the web.
Managing materials for audit purposes.
Improving efficiency ... but entirely focused on management and admin issues. Nothing about the work of individual researchers yet.
Data driven research
Sharing datasets.
Claim that repositories allow seamless navigation between different data sources - I beg to disagree, it's the surorunding infrastructure that might achieve this.
Life-long learning
Underplayed aspect?
Adaptation of teaching materials.
Ummm... what's this to do with lifelong learning? It seems more like "lifelong teaching" issue.
This leads me to thinking about a model that does not separate the work/learn activities. People learn by doing. Work is doing. Can e-learning techniques be adapted to supporting knowledge workers in their work? Maybe this way, a new strand of personal and company revenue can be tapped. Professional societies do some of this, but ultimately there's no tutoring element. Maybe something like a distance research degree, based on work. (Who was the person talking at the Edinburgh stargazing group about research first degrees? It was Kevin Thompson, Anglia Ruskin University, brief notes or his talk [here])
Preservation
Slides states the obvious. Talks about open access, peer review, etc.
Repositories are central, but not the whole answer.
Need for user-driven approach, not technology -- but if tech is enabling new working practices, can't just ask users what they want.
Questions
Les Carr: notes that the talk clearly articulates how the open access debate has moved from narrow concerns to general university policy areas - e.g. branding and marketing.
David: impressed by emphasis on data, but current repositories focus on papers. A: yes, I do want to generalize, not create another battleground. Need to be broad-thinking.
Chris Rusbridge, Edinburgh. Question about cessation of AHDS funding. A: notes that underlying problem of unanticipated cut in AHRC(?) funding.
Howard Chase, University of Cambridge: should e-learning materials be available OA. A: abstractly, ideally, yes, but there is a context and business concerns. Maybe, need to separate funding of materials from funding of teaching?
Data breakout
David Shotton - Biological Images
- Expect presentation to appear [here].
Defining Image Access ... mandated to look at images accross subject boundaries. Outputs: report and wiki.
Main problem with images - they are not self-describing - metadata is needed to expose the semantics (to machines).
Publications as rhetoric. Selection of facts to support rhetoric. Most data never published. Historically this was necessary. Things are changing with online publications. But this capability is not widely used.
Take home:
- datasets as 1st class publication objects
- domain metadata
- papers/data cross-linking and interoperability.
And so to data webs...
And Image Web project ...
A data web for images.
Desire to integrate and cross-search image repositories.
Repurposing: e.g. evidence for cancer therapy vs teaching examplar of confocal microscopy.
Stakeholder analysis: see slide. Academic publishers ... copyright ... s/normal/appropriate/?
Making the vision ...
Requirements analysis.
What we did (see slides):
- talk to repository holders
- evaluating software
- evaluated a range of software tools
- advanced ideas and concepts
Achievements (see slides)
Conclusions - some of these are not new conclusions
Future plans ...
Interactions ... other's (sic)
Anne Robertson, Edinburgh
Scoping Geospatial repository - GRADE.
June 2005 - April 2007.
National Ocenographic Centre Southampton, Pauline Simpson
Examples of geospatial data ... (I prefer Mr Grizzle)
Digimap. Research is both a consumer and producer oif geospatial data. Sharing should be more actively supported.
Work packages:
- digital rights
- informal mechanisms
- user based evidence
- role of institutional repositories
Legal issues
- copyright not in geospatial data, rather use EU database law
- means substantial copying and redistribution of part, even substantial, is allowed without permission or attribution.
Sharing
Informal sharing:
- data is shared informally, frequently
- use CD or email attachment
- finding: Google, Edina services, US clearinghouses
- barriers to sharing: licenses, lack of quality metadata, protecting own IPR
- desiderata: more permissive licensing, national repository
More on current sharing preactices:
- technical means not well deployed
- dependent on people networks
- many datasets that could be made available for re-use (900 at 4 sites)
Demonstrator
Chose Dspace ... because of contact with another user. Literature review did not yield much advice.
Findings about searching. DC sufficient to assess fitness of data for investigation, with extent of data thumbnail.
Desiderata:
- Download: direct to desktop, full human readable geospatial metadata
- Upload: packaged data, would provide geospat metyadata, auto creation of bounding region, assist with licensing statement
Institutional repositories
Most institutional repositories focus on papers only. Data is not fully supported.
Where designated data centres exist, IRs probably not repository of choice
SWOT analysis - see slides
Summary
Repositories have a role for geospatial data
Lots of informat sharing
Community would like means to share
Barriers:
- licensing
- metdata
- protecting own IPR
Institutional repos not suppoirting geo data
Geo data creators support reuse, but there are difficulties.
Question
How confident about legal opinion? This was just one specialist lawyer's opinion. But similar opinion has been expressed separately w.r.t. US law.
Images - panel
Discussion about problems with re-use of images. CLA? Heron? e-Content alliance?
Funding puts people in competition, but we want to help people work together.
Mike Pringle: If we go to the JISC with this problem, they are likely to ask "what do you want us to build"? This is not helpful.
Common infrastructure ... now understand what is wanted? Not quite ... rather a framework within which the usage model for each contribution is clear to all users.
A take-home message here is that the IPR issues are probably the major impediment to re-use of images.
Mike: another message - start talking about what is made possible when the problems have been solved - emphasize the benefits to all parties. Maybe JISC can fund exemplars of what becomes possible when these problems are solved.
Problem of having lots of storage but not able to find stored materials.
Final plenary
Overview
Reviewing the roadmap for greater sharing and re-use
Policies
Open access mandates - access and preservation. Also good information management. More than Flickr can do.
- OpenDOAR policy tools
- OA is not self-evident and not enough?
- reward structures
- StORe: professional development, changing role for data libraries
- Web 2.0 as DNA "ladder" - self organizing, self-replicating, ... hmmm
- Legal framework
- employ as enabler rather than constrainer
- licence registry
- licence to publish, retain ownership
Technical development
- Better object modelling
- DC Schol app profile
- other app profiles
- OAI-ORE
- Better linking
- StORe and CLADDIER
- Rich tags
- what about persisting links?
- Better interfaces
- Interoperability demonstrators (Soton, Cambridge-Jim Downing)
- Access management
- Shibboleap ?
- ASK and BID
Without metadata, can we be better than Google? Can we manage? Can we cite?
SOA is too hard ... (depends on what you mean by SOA)
Creating a repository network:
- Repositories research team
- Repositories support team
- The depot
- Intute repository search
Panel
Fred ?
Roadmap and milestone:
- no single approach will get us there
- open access is not enough (means not end)
- end is well-being of academic process (research, teaching, learning ... no mention of transfer/exploitation).
But we have made great progress.
Sheila Anderson
(Director AHDS)
Praise for the JISC.
AHRC reasons for decision to defund AHDSD, given general agreement of needs:
- community has sufficient computer literacy to create and manage dig resource
- HE are in position to take on responsibility
In light of this, response:
- They are throwing responsibility back to the community. JISC capitial program won't continue. Who pays long term for infrastructure, and to enable long term plans? Thye principle of funding was previously accepted for libraries, but not yet not for digital content.
- Too many library/computing people here; not enough researchers. Repository program is great, but so far it's a failure.
- Take more risks. Innovative projects. More of this. More responsive mode calls. Build on trust. Empasize information, content, relationships.
Rachel Bruce
Conference program shows learning, teaching as well as research is being taken seriously.
Questions:
- incentives to share learning materials?
- global services vs institutional services
- metadata vs tagging
- are we truly addressing barriers to OA?
- how to be more directive(?) to build network of repositories.
I've noticed many interpretations of Andy Powell's plenary - it would be nice if he had chance to respond in the final session.
Floor questions
Should there be mandates for depositing e-theses?
What to do about IR holdings in AHDS? (AHDS being recently defun ded in a surprise move).
Concerning ease of submission - simplicity is not enough, still need paid curators to organize and manage materials.
Peter Murray Rust - reactive call for proposals from students for collaborative research systems?
To encourage thesis deposition: external examiners might ask for URL rather than copy.
Need more vision of what could be: what we do is worthy but dull. (Next question from "nice but dim" - Peter Burnhill)
In the wake of the AHDS defunding decision, which was unforeseen and is indicative of problems in policy formulation process: how is the gap in this area to be filled? How isbthe required skill-set that was AHDS to be alternatively deployed.
Rachel: JISC is not infrastructure provider for RCs - RCs are infrastructure providers (except in certain parts).
Keith Jeffery - a question remains unanswered: where is primary responsibility for repository infrastructure provision. Also want multiple copies, notwithstanding primary responsibility.
