Science Library Pad

Fedora repository conference in Wales October 24, 2005

If you've been reading about the Euro Fedora [repository] meeting and wished you could be there, luck has it that the National Library of Wales is also hosting a meeting next month, and although registration has closed perhaps the nice chaps there may still let you in :)

It's called Digital Asset Management with Fedora [repository].

To be held on the 24th October 2005

Hosted by Llyfrgell Genedlaethol Cymru / The National Library of Wales, Aberystwyth, Wales, UK

Euro Fedora User Meeting 2005 - presentations are up

Most of the presentations from the European Fedora [repository] User Meeting 2005 are now up.
They are linked as e.g. (ppt) within the program.

Also, as I mentioned previously, most of the presentations from Building the Info Grid 2005 are up as well, on the speakers and abstracts page.

if you're blogging at Internet Librarian 2005 (IL2005) USA

Nancy Garman writes in the InfoToday Blog

Watch this space for news about Internet Librarian 2005, coming up next month in Monterey, CA – October 24-26! The ITI editors and staff will be blogging the conference live, plus we'll be sharing pre-conference tips and news in the weeks before the conference.

If you’re a blogger and plan to attend Internet Librarian, drop me [Nancy Garman] an email [[email protected]]. We’ll post a header on this blog with links to all the IL blog coverage, creating a Blog Central for Internet Librarian 2005 [USA].

I think that's a great idea.  I myself won't be there as I'm going to the London conference, where I will be blogging in some fashion.  After blogging two conferences this week, I plan to have some musings about conference blogging in general coming up as well.

Euro Fedora User Meeting 2005 - Fedora as an OAI-PMH (and WS) Compliant Data Provider

Fedora as an OAI-PMH (and WS) Compliant Data Provider (PowerPoint)
Ana Macario, Alfred Wegener Institute

* very data-centric organization
* involved in DataGrid

* SOA at AWI
[diagram of 2004]

In practice...
post-print -> repository
PI is supposed to release data when published, but by then, it is lost or there are excuses

So need Staging -> Publication

So: Fedora to do all this.

* Reasons for Fedora
- Virtual Repository
- not restricted to Dublin Core
- standards compliant
- etc.

[diagram of 2005]

currently using dc.source and dc.relation as a hack to express linkages,
convinced that proper way to do this is RDF/XML

Long-term issues
* Benchmarking for large number of files
* out-of-box web client acceptance
* fine-grained access control and Shibboleth based AuthN - relevant in DataGRID
* support for sets
* federation model
* collaboration and support
- disseminators for visualizations services; relevant for DataGrid
- Eclipse project to facilitate plugin devel
- Google strategy?
- seminars, tutorials for advanced Fedora users

Euro Fedora User Meeting 2005 - Introducing Pergamos: A Fedora-based Digital Library System utilizing Digital Object Prototypes

Introducing Pergamos: A Fedora-based Digital Library System utilizing Digital Object Prototypes (PowerPoint)
Kostas Saidis

- provide the means to treat content variations in a unified manner

- for digital object they want behaviors according to nature: if it is a book, give us the pages, we
don't care about the encoding

Digital Object Typing Info

* Humans interpret Content Models (in this case)
* problems: tool weaknesses

Preferred: DL should resolve DO type automatically, e.g. force conformity

DO Object Prototypes, based on OO methodology
- content
- private behaviors
- public behavior

Euro Fedora User Meeting 2005 - Fedora Content Models for the National Science Digital Library Data Repository (NDR)

13:15

Fedora Content Models for the National Science Digital Library Data Repository (NDR) (PowerPoint)
Carl Lagoze
Cornell Information Science

... Information Network Overlay

a digital library (particularly science DL) is more than just documents: it's data, orgs, people, services

all (primary resource) objects named with handles in overlay

... including Access Controlled API

interested in very rich granualarity and access control

some of this is still theoretical, not in production yet

1 million metadata records in Oracle (Dublin Core) will = a digital object in Fedora
there are ancillary objects around the metadata = about 2.1 million objects

NSDL is not a content library, it is a metadata library
but it will be possible to add content e.g. lesson plans

each object turns into about 60 triples = about 60 million? 120 million triples

NDR currently one Fedora repository, will become ? federated repositories

want to build representations of people in the NSDL, so that you can build a reputation system for those people
"this person is expert in X"

can use proai to expose using oai anything that you can connect with a disseminator?

make sets using aggregation

Euro Fedora User Meeting 2005 - Researching Fedora to Serve as Central Repository System of the State and University Library

Researching Fedora to Serve as Central Repository System of the State and University Library
Stephan Drescher (PowerPoint)
and Birte Christensen-Dalsgaard (PowerPoint)

* National Library - digital preservation
* Storage Preservation
- as many different systems as different objects (currently)
- including both digitized and born-digital (IR and Webarchive)

working on automatic ingest
working on national infrastructure for institutional repositories, with infrastructure in charge of
preservation - national infrastructure for storage with redundancy etc.

Preferred preservation strategy: migration

working with Royal Library in Holland who have an emulation program running with IBM

Stephan: Archiving of Denmark's Broadcasting of Radio and Television (BART)

* covering 24/7/365
* 220 GB a day
* data needs to be evaluated and eventually corrected after 48 hours
* automatically ingested into repository

How to fit Fedora within Bart's resource workflow

Uses Linux and off-the-shelf.

80-100 TB a year

Q: are you going to put this data into the repository?
A: into the repository only references (to the raw data) will be input

Euro Fedora User Meeting 2005 - DiPP Digital Peer Publishing

Digital Peer Publishing
DiPP Fedora-based system for Open Access eJournals (PowerPoint)
Jochen Schirrwagen
hbz

http://www.hbz-nrw.de/
http://www.dipp.nrw.de/

Mission
- foundation of new and expansion of existing scientific electronic journals
- customizable workflow within a common publication system [including peer review]
- fast, open and transparent digital peer publishing

Components and Middleware
- Peer Review System (external component)
- from Peer Review System via LDAP auth to Plone-based Publication System
- from Peer Review System via OAI to DiPP Services Fedora-based Repository
- from DiPP to Publishing system via SOAP

currently using Fedora 1.2

challenge: external peer review, Plone-based publishing
want: at least appearance of one unified system

Publication Pipeline: includes Repository Engine, Conversion Engine, Publication Engine
therefore DiPP Services
* Conversion Service
- uses commercial tool to get to DocBook XML, then XSLT to transform to HTML and even PDF
(convert XML to LaTeX using XSLT then to PDF?)
* URN service
- persistent identifier for journals, articles, supplementary material
* Distribution Service
- registration for OAI harvesting
- RSS feeds
- email alerts

also had to build Hierarchy Service because of limitations in Fedora version 1

Fedora - is it good enough?

pros
- highly modularized
- versioning of datastreams
- addition of own metadata formats

cons
- added metadata in own formats not searchable
- inconsistent versioning of API-A (Access) and API-M (Management) from Fedora 1 to Fedora 2

would be good to have a pool of common services

Summary and Outlook
- workflow-based publication system
- extended services on top of Fedora

Next Steps
- DINI certification (German centre for network information)
- full-text indexing via FAST search engine
- improvement of usability

[email protected]

Euro Fedora User Meeting 2005 - Institutional Repository In A Box

Institutional Repository In A Box (PDF)
Christian Tønsberg

DTU uses local (specialized) version of National Research Database

Current state:
- Data Production App: MetaToo
- Storage
- Indexer (zebra)
- Web Front (HTTP, java/tapestry, sru/srw)

want to talk to Fedora instead of National Research Database software
- generalize Data Production Application
- generalize Web Front
- should work out of the box

CVT Project Fedora2 (IR-in-a-box)
http://defxws.cvt.dk/projects/fedora2/

Euro Fedora User Meeting 2005 - 09:20 - Fedora Project Update

Fedora Project Update (PowerPoint)
Sandy Payette

version 2.1 of Fedora embodies the full core system they had envisioned

Fedora for Digital Archives and Records Management

interest in RDF capabilities

designed so that modules can be replaced

improved OAI provider service coming in 2.1, can also be used separately

objects stored as FOXML (Fedora Object XML)

variety of ingest/export formats
- FOXML, METS, METS 1.4, MPEG21 DIDL

Fedora 2.1
- introduce Fedora Service Framework (services distributed as part of the official distribution)
guarantee compatibility as Fedora evolves
- intro PROAI (OAI Provider Service) ? on sourceforge now ?
- intro directory ingest service

Authentication and Authorization improvements

* AuthN plugins
- HTTP basic auth
- Tomcat realms and login modules

* SSL

* AuthZ
- XML-based policies using XACML

developing a user interface for policy building

RDF-based Resource Index (RI)

* ontology of common relationships
* stored in Resource Index
- automatic index (using Kowari triple-store)

* RI search (search the repository as a graph)

* new in Fedora 2.1 for Resource Index
- scale and performance testing (NSDL 2 million objects, > 100 million triples)

RDF relaxed rules now permits relating to objects outside a specific repository.

There is a plan for federation and naming services.

improved logging using log4j

handle system plugin for PID generation

you can rebuiild the entire repository by crawling the FOXML,
there is now a Rebuild Utility for Repository Indices

FedoraClient utility class for building new (SOAP) clients

Fedora Future

2006
- better interface: FIRE client
- BPEL workflow engine
- advanced Fedora search
- preservation integrity
2007
- federation pid resolution
- preservation monitoring
- event notification (pub/sub)

Pathways InterDisseminator
- OpenURL access point

- "content model" specification language
- advanced object creation workbenches
- Shibboleth and Web Services security

Community Working Groups

- Preservation (Rutgers)
- Workflow (Peter Murray, OhioLink)
- Outreach (Rutgers)
  = improve Fedora web site
  = collaboration environment - Wiki, Confluence, other?
- Content Model Working Group (under charter)

There is a Fedora Advisory Board.
- define sustainability model

* there are collaborative development opportunties

* will put up a page on http://www.fedora.info/ for user-contributed tools

Info Grid 2005 - Tuesday 27th, 17:00 -

17:00

DEFF Identity and rights mangement activities
Arne Sørensen, DEFF

[err, see the presentation, I'm tired]

Presentation: ppt (466K); pdf (1.4M);

Projects 2005

* Shibboleth and EZproxy

http://www.deff.dk/aai/ (currently points to http://www.statsbiblioteket.dk/AAI/aai.jsp )
http://sso.kb.dk/

Projects 2006
- international networking
- use Knowledge Exchange
- national workshops
- from devel to production in selected areas
- formation of Danish InCommon: DK-AAI

Organization issues
- formal org to guarantee the sphere of trust
- e.g. Haka in Finland

Projects 2006+

- Shibbolize user databases
- Shib - ScienceDirect trial

Info Grid 2005 - Tuesday 27th, 16:20 - JISC Middleware Programme and Shibboleth

4:20
JISC Middleware Programme and Shibboleth
Terry Morrow, JISC Consultant

[background]

JISC Core Middleware
- authentication
- authorization
- directory services
- identifiers

[stuff about Athens - centralised authentication + authZ - technology + infrastructure (people)]

Issues:
* Athens requires management of separate "Athens accounts"
- Recent development (AthensDA) uses local IDs
* Little take-up of Athens outside UK
* Service providers have to license Athens
* Centralised service - relatively high operational costs
* Not well suited to increasingly complex authN scenarios
* Meanwhile other countries moving to SAML/Shib

JISC Core Middleware
- tech devel
- infrastructure

Aims:
- better understanind of middleware
- build working Shib infra
- support takeup of Shib

http://www.jisc.ac.uk/programmme_middleware.html

* programme has funded 15 different projects so far

tech investigated
* PERMIS
* RADIUS
* Shib

Core Middleware - Infrastructure

Aim: establish a working Shib infra

work areas
- Data Centre services (MIMAS and EDINA) to be made Shib compat
- create Athens-Shib gateways
- funding early Shib adopters
- creating service to assist adopt
- estab national UK fed (to be known as *Sparta*)
- liase with suppliers/vendors

Eduserv has written their own IdP

UK Data Archive - SAFARI
- Access control to a wide range of social science survey data

Future

- the UK Federation - what to do
- UK WAYF will be established

Challenges
- cultural change: shifting functions from libraries to computing science

Info Grid 2005 - Tuesday 27th, 15:35 - Access Management projects at LSE (PERSEUS and ShibboLEAP)

15:35

Access Management projects at LSE (PERSEUS and ShibboLEAP)
John Paschoud, LSE (London School of Economics)

an institutional (university library) perspective on cost, benefits, etc.

[some history]

Athens
- "a big database table with 3 million rows and 300 columns" (somewhat facetiously)

PERSEUS
- LSE Shib-IdP had been previously established
- used to explore
  = access to end-user info systems
  = [other things I missed]

PERSEUS relationships

* used to bind 4 different internal systems, and then with others: uportal, Internet2, JISC, Endeavor,
  WebCT, SITS...
* relationships with other projects:
  DART (Digital Anthropology Resources for Teaching),
  nereus (sharing economics resources), ShibboLEAP

Federations?

LSE is in lots of federations

Terms of content licenses

* Initial scoping audit of LSE Library e-resources management and current e-licensing situation
* Active participation in Meridian (Endeavor e-resources management)
* Active particpation in major e-licensing initiatives
* Joint PERSEUS / NEREUS study of e-resources Access Management terms (across 6 Euro countries)

Looked at roles of users

[list of roles]

Some challenges

* slow progress with Signet/Grouper, uPortal, and WebCT
* need for a central UK (or wider) database of standard licenses for major suppliers that reflects
  Shib-usable (eduPerson) attribs

Alumni groups as Virtual Organisations

[list, challenges]

ShibboLEAP

* get the UK Shibboleth ball rolling, something about access to Eprints repository?  ~150k people
* Role-based access in open archives: who is permitted to do what

* creates IdPs sitting on existing infrastructures at the organizations
* enables Eprints as Shib SP

Project Management

* Regular Library AND IT staff at each site
* High-level buy-in
* Focussed Project Management Board
- defined tasks for each meeting (one thing for them to decide)

Shibbing Eprints

* AuthN: easy
* AuthZ: not so easy

Benefits include
* other institutions can use this as the basis for their Shib IdP projects

a bunch of URLs including

http://www.lse.ac.uk/library

http://www.angel.ac.uk/PERSEUS

http://www.angel.ac.uk/ShibboLEAP

Info Grid 2005 - Tuesday 27th, 2:25 - Shibboleth and ScienceDirect

2:25
Shibboleth adoption by content providers (Shibboleth and ScienceDirect)
Ale de Vries, Elsevier Product Manager, Science Direct

[Elsevier marketing stuff...]

[background on authentication]

Federated Authentication

Pros:
* Allows access to remote services using campus login credentials
* No extra admin overhead for customer
* Simpler for users
Cons:
* Complex to implement
* Requires agreement on "rules of the game" between all vendors and institutions

Elsevier's view

* Shared / Federated
- fulfills customers needs
- win/win (less admin)
* we will continue to anonymous, campus-wide access whatever the tech
* we will continue to offer PZN in exchange for basic end-user registration

Shib benefits as Elsevier sees them

* replacement for IP authN
- removed admin burden of IP maint
- removes dependencies on network arch
* allows PZN based on local creds
* removes need to remember multiple user/pass
* avoids problems of proxy servers
* helps us provide the broadest possible access to our customers

Shib and SD: ramp-up

* April 2002 - workshop
* May 2004 - Shib release
* based on Shib 1.1
* held workshops to involve customers and Internet2
* Findings
- anon non-personal a must
- provide option to PZN with opaque unique ID
- needed support for deep linking (need to authenticate no matter where on the site the user enters)

Shib and SD: ... testing...

* Dartmouth, Georgetown, NYU, UCSD, Penn State
* no major problems
* none of the pilot participants rolled out access to broad user community

Shib and SD: production

* Fed 2005: Moved to InCommon
* July 2005: multi-federation support due to release
- main issue is branding and IdP discovery in a multi-federation world

Current Status

* University of Southern California: up and running in May
* Working with SUNY Buffalo and Georgetown
* in discussion with JISC (UK), SWITCHaai (CH - Swiss), SURFnet (NL)

WAYF issue

- multifederation: no one runs a WAYF of WAYFs

:( end users don't understand federation concept
:) but federations are geographically oriented

[example of Shib login to SD - recorded live in production]

uses a cookie to remember your organizational affiliation

Open Issues

* Tech new, complex and rapidly changing
* Federations are in very early stages
* Uptake is key... we are at a critical point
* need to make implementation easier for smaller customers and vendors
* Elsevier is committed to making access easier for users and will continue to support Shibboleth

Tech: [email protected]

Q (from AA of Google Scholar): do you see a lot of IP spoofing?
A: no

Q: assuming that you are planning to offer Web Service access to Science Direct... do you have plans
to offer access control
A: Web Services... well... not yet.

Q: what do you log? user path through site?
A: yes we log, analysis is on aggregated basis, we have to log path to do reporting to customers,
and for product improvement

Q: to what detail do you know the user when they login using InCommon - e.g. student from university X?
A:
1. can come in using provider ID - identifies institution
2. can come in using target ID - identifies individual - dependent on federation - if not wanted for
personalization, Science Direct discards this ID

Q: do you see a lot of use of personalization in your product?
A: quite some - between 5 to 25%

Info Grid 2005 - Tuesday 27th, 13:30 - Internet2 and Shibboleth

13:30

Internet2 and Shibboleth (Shibboleth and Privacy)
Peter Brantley, California Digital Library

Presentation: ppt (0.8M); pdf (1.3M);

Access Management

list of requirements from A White Paper on Authentication and AccessM anagement Issues
by Clifford Lynch, CNI, 1998

Distributed Access Management

registration and authentication: library, college, university
authorization and accounting: the resource owner

Role vs. Identity

Conceptually, licenses to restricted content are by ROLE e.g. {member of} University of California

Therefore IP address authentication is inherently kludgey.

Improved AM

* something opaque that a service can use to associate with me and give me access
* if a service provider knows where you are from, it can ask your identity provider
* once authenticated, access can be determined by role
* SP can ask therefore only for relevant attributes
* might be {member of}, or {staff}, or {IT Director}
* this approach enables finer role-based distinctions

This is Shibboleth

three things:
* Shibboleth Project
- umbrella of activities around federated authentication and access management, still ad hoc
* Shibboleth Specs
- SAML 1.1 and enhancements
* Shibboleth System
- Internet2-developed open source (reference) implementation
- there are other implementations available

Shibboleth status
- Elsevier Science Direct, OCLC, JSTOR, EBSCO, Proquest, Blackwell
- international uptake: Switzerland, Finland, UK, Australia
- production federations (e.g. InCommon)
- discussions involving "Leagues of Federations"
- supports US Federal E-Authentication Initiative
http://www.cio.gov/eauthentication/

typically will send both institutional attribs (e.g. contract #) as well as individual attribs

could have an attribute release policy interface so that users can control who gets what attributes

Shibboleth Federations
- Usually nationally-oriented
- Federation operator handles coordination and managment of operations
- coordination, management, sets of atrribute-types, what to trust, for what purposes, policy, operational direction
- Federations are very hard work, not to be taken lightly

Meta-Federations

- Shibboleth InCommon and US Federal Federation (FedFed) are establishing peering
- Sun's Project Liberty and Shibboleth will interoperate through advancement of technical standards
- increasing need to support Virtual Organizations (VOs)
- expect eventual formation of Shibboleth-enabled software-based infrastructures, extending the concepts of
VOs, that support personal attribute requests, very small groups

[Ohio State is one of leading adopters]

Meta-Fed Discussions: US, UK, Netherlands, Finland, Switzerland, Australia, Spain, ...

Issues:
- policy framework
- privacy needs
- working with multinational service providers
- scaling WAYF

Slaughter meeting, Oct. 2004
Good meeting, not wholly conclusive.

Different authentication technologies and infrastructures exist within Europe.

EC funded GN2: eduroam, eduGAIN

InCommon...
operations

Origins and targets establish trust bilaterally in out-of-band or no-band arrangements (using shared
posting of practices)

InCommon progress

- Eduperson attributes

more challenging:
- having apps make intelligent use of federated identity
- legal issues
- scalable paths for LOA [?] components

Shib communities

Flexible creation of federations within federations, of varying duration, are increasingly common
in the US
- RedCross for Hurricane Katrina
- UCTrust

Next Steps
- multi-Federation world
- enhancing support for US gov auth
- support GridShib
- encourage continued commercial and vendor uptake
- interop with Microsoft
- extending Shib beyond the browser, into service-service interaction
- Shib 2.0 utilizing SAML 2.0, enhances functionality.  Design has been committed.

So... what's SAML?

[explanation of SAML]

An XML-based framework for exchanging security information.

Shibboleth and SAML

- Shibboleth is a profile of SAML
- extends SAML through Shib's data flow specs

Privacy and User Consent

- idea is that you only release what attribs you specify

Trade-offs

Anonymous use hinders PZN; persistent anonymized identifiers permit a range of PZN options.

* More dialogue on privacy and policy best practices is necessary.

Comments: JSTOR - little Shib uptake - used as a research project - everyone is using IP
following Q: why isn't there more adoption?  what is needed for broad-based adoption?
A: libraries say "what we works well enough" using IPs / proxies / VPNs
there is also a transition cost, user education

Shibboleth is currently gaining more tracking in accessing administrative apps (e.g. financial benefits management).

not much traction in university libraries

However, there is no better system available for this type of role-based authorization.

[Q about US PATRIOT Act and privacy implications]

Info Grid 2005 - Tuesday 27th, 13:20 - welcome to identity and rights management theme

13:20

Second theme: Identity and rights management

AAI - the scene

Authentication, Authorization, Infrastructure

Defining and organizing a common framework for identity and rights management administration within and between organizations
for (higher) education, research and their e-resources.

The Vision

Users at one institution can access external resources without needing to be a registered user of the external organization.

Rights and roles can be set in Attribute Release Policy (ARP).

The resource provider collects attributes from the user home
organization WITHOUT KNOWING THE IDENTITY OF THE USER.

lots of advantages, including economic

What is not part of the framework

* What material that may be accessed and by whom
* Payments or payment mechanisms
* Agreements on purchase of material ...

Info Grid 2005 - Tuesday 27th, 11:40 - Overview of DEFF SOA activities

11:40
Overview of DEFF SOA activities
Mogens Sanfaer, Technical University of Denmark

they use Web Services interchangably with SOA

[diagram of 3-tier DEFF architecture 2000-]

* one common service infrastructure - many portals
* facilitating cooperation as well as competition
(services can compete)
* paving the way for a re-integration of library and otehr university infrastructure domains
e- publishing / libraries / research / learning / governance

[diagram about integration promise of web services]

[diagram that I don't understand about the b2b nature of the service landscape]

is it possible to distill common services out of current monolithic systems?

Implementation Roadmap
1. implement pilot web services
2. facilitate pilot exploitation
3. implement more web services
4. establish cross-domain scenarios
5. contribute to open source tools and international cooperation
6. bring vendors into the game

1. Pilot Web Services
- global eprints using Fedora
- digitised Danish journals based on Oracle
- XML to Z39.50 gateway ->

HTTP SOAP/REST Web Services gateway is open source, available for download

[diagram]

2. exploitation project Web Services Testbed (started this year)
- training courses, workshops, consultation
- explore technical means to ease exploitation such as WSRP, UDDI
- gather feedback - trying to find the killer app
- discuss business models issues

it may take a bit of time for library uptake of Web Services and SOA concepts and technologies

3. more Web Services

pull from DEFF e-learning

journals, books, everything as Web Services - for consumption by Blackboard, CampusNet, Open Source LMS...

*** projects trying to satisfy these needs

4. beyond the library domain

DEFF e-publishing is building a national research database and repository architecture based on XML Web Services

* The collaboration with the Danish Universities Digital Governance Project has created an opportunity
to integrate the admin and library domains e.g. systems to create reading lists

* would also like to be able to create a national portfolio system for students, to capture their work

(my note: another application that has been discussed for these systems is national research portfolio for scientists)

["complex" architecture diagram]

* Is there a simple architecture?

synchronize local student portfolios (using harvesting and aggregation) with national portfolio
requires standard person id

how do you scale to e.g. pan-European? intercontinental?

5. contribute to open source and international cooperation

* X2Z gateway released, more to follow
* Fedora as generic repository architecture
- contribute especially in the areas of search and preservation services
- Euro Fedora User Meeting

6. bring vendors into the game

* Library system vendors et al. are starting to show interest in Web Services

Perspectives for the next couple of years

* implementing more web services
- OAI-based cross-searching services for research databases and library catalogues
* facilitating more exploitation
- implement best-practice examples of portals based on Web Services and open-source software
* crossing library domain boundaries
- contribute to the development of e-Science infrastructures in Denmark
* open source
- continue Fedora cooperation
* bring in commercial vendors
* exploit VIEWS - Vendor Initiative for Enabling Web Services

Info Grid 2005 - Tuesday 27th, 11:00 - The DARE architecture and SOA-developments in the Netherlands

11:00
The DARE architecture and SOA-developments in the Netherlands
Martin Feijen, SURF

Outline
* SOA in higher education
* DARE technical architecture
* DARE organization
* conclusions and questions

help libraries to make the changes needed to support scientists in their work

SOA in higher education

- SURFnet (people over 18) is planning to use SOA in the very near future
- awareness of performance issues: protocol overhead may slow down inter-application response
- SURFnet Video Portal and DARE harvester are the first SOA implementations within SURFnet

SURFnet is making the case to their management to do SOA, but the management says "what is the business case"

Kennisnet (Knowledge net) - people under 18

- implicit SOA policy: focus on re-use of applications
- SOAP/XML interface between local web apps and central portal
* architecture is SOA based but
- to avoid performance, constraints are necessary
- use only when applicable, not blindfolded

SURF
- Task force group "Information Architecture in Higher Education" April 2005.
  No specific recommendations about using SOA.
- Report of the Scientific Council of SURF 2007-2010

  Advice: use SOA as architectural framework for the further development of the technical infrastructure.

DARE architecture

- building a network of institutional repositories in the Netherlands
- uses OAI-PMH
- DARE is successful but we have a need for optimization

[diagram of DARE as is]

Limitations
- app interface between local repository and other tools (e.g. Metis system)
  = organizational issue
- no consensus persistent identifiers
- no solution for complex documents
- no unique identifier for personal names
- metadata quality issues
- sets and/or filters

Planned work on Technical Optimization

* content spider and filter (want to include datasets and not only publications)
* Digital Author Identifier (each scientist in the Netherlands will be assigned a number) - will integrate with OCLC/Pikas
* pilot for e-theses using DIDL and extended OAI
* persistent identifier

LOREnet project

* demonstration project to create 15 operational learning object repositories, building on DARE architecture
* OAI-PMH, LOM and IMS-CP as standards

[diagram of Lorenet]

DARE organization

* Libraries need to change
* We want scientists to use repositories, but we don't speak their language
(Approach has been inside out - from library to scientist.  Need to change to outside in.)

[diagram of organization, showing missing links between faculty and library]

DARE approach

* http://www.creamofscience.org/
* library self-evaluation and summer school for libraries
* embedding DARE in library workflow
* get closer to faculty and students
* not perfect but "good enough": light weight, flexible and little steps (SOA might be very handy to support this)
* need to do architecture AND manage change
* facilitate and encourage

DARE-II (2007-2010)

* Primary goal: repository as a tool for research and learning
* collect, store, describe, disseminate and secure all digital objects that are relevant for scientific communication
* So: not only end products like PDFs or articles but also datafiles, models, learning materials etc.

Conclusions

* SOA is known mainly in the ICT community
* SURF, SURFnet and Kennisnet will use SOA (prudently)
* DARE will migrate to DIDL and extended OAI infrastrcuture
* DARE will facilitate libraries in their change process and to go beyond publications

http://www.surf.nl/

Q: why DIDL?
A: have been talking about compound docs for about 2 years, had meeting to look at solutions...
decision was DIDL

Q: if you have persistent identifiers, won't you need a resolution service - one exists - ??Swedish steff program???

[more questions]

Danish Royal Library - Black Diamond - photo

  IMG_7407.jpg 
  Originally uploaded by rakerman.

The "Black Diamond" (Den Sorte Diamant) extension of the Danish Royal Library (Det Kongelige Bibliotek).

Info Grid 2005 - Tuesday 27th, 09:45 - Rebuilding Information Services for the Digital Age

9:45
Rebuilding Information Services for the Digital Age
Norbert Lossau
Bielefeld University, Germany

Presentation: ppt (2.9M); pdf (3.4M);

what does the information environment look like for implementing SOA

Key questions
1. How do SOA-based activities relate to our existing information environment?
2. "Collaboration between different communities is needed" - but how do we organize this collaboration?

Information services provided by libraries
* search
* access
* re-use
* communication/dissemination/publishing

New developments
- grid (mainly middleware)
- e-science

End-users demands grow and differentiate
* researches etc. creating compound, live documents, changing nature of knowledge generation
* more advanced searching and browsing e.g. integrate search into desktop tools
* better access to more things - in an interdisciplinary environment, people need access to more information
* re-use
* communication/dissemination/publishing

A Conclusion

* Libraries cannot possibly cover all these demands

The need for rebuilding information services

* We already have too many systems with partly overlapping functionalities, often not interoperating with each other

- leads to confusion on end-user side
- leads to inefficient deployment of staff and financial resources

SOA opens up new ways to *re-use* information services and promises *interoperability* between heterogenous systems and [something I didn't get]

[diagram of service modules of Digital Library NRC at Bielefeld University]

the system is a set of services

they are working on replacing existing metasearch service with BASE/Google Scholar/Scirus/ ...

build on state-of-the-art technology - IF it's possible to integrate them

Potential

* opens up unrealized opportunities for collaboration

... joint rebuilding of information services based on SOA?

SOA is an opportunity to restructure how we work together.

no explicit SOA activities in the German library world

http://www.dl-forum.de/

* in the German library world, SOA mainly understood as Web Services, implementations have only been slowly progressing

There have been digital library developments, however.

http://rzblxl.uni-regensburg.de/ezeit/?lang=en

http://elib.uni-stuttgart.de/opus/doku/english/index_english.php

http://www.redi-bw.de/

BASE (Bielefeld Academic Search Engine)
http://www.base-search.net/
* based on FAST data search
* Web Service interface is planned

http://www.hbz-nrw.de/produkte_dienstl/DigiBib/
* internal Web Services
* external Web Services planned

Digital Peer Publishing (DIPP)
http://www.dipp.nrw.de/
* Publishing workflow and online publication system for OA journals
* Tech: Fedora, Plone, GAP - internal SOA, no external Web Services

Outside the German library world, there are more sophisticated technology developments, Java based.

development of Lucene for metasearch - L3S, Hannover
http://www.l3s.de/english/news/about_the_L3S.html

eSciDoc
http://www.escidoc-project.de/startseite.html
- buidling an integrated information, communication and publication platform for network based scholarly work based on Max Planck Society
- SOA architecture using Web Services
- scholarly workbench, publication management, etc.

Outlook: can we build inventories of information services - "yellow pages"
- regional, national, international

Requirements:
* create business models
* define basic standards
* provide support and maintenance

In this environment, how do we define a *product* that we deliver, and how do we measure the success of the product?

Define success:
1. service developed?
2. service developed and running in production?
3. service developed, and used and re-used in production at many different organizations?
4. #3 plus wide acceptance by many users?

Q: what could we do to help services move into real production?
A: we have a development department (in the research community) but no sales and marketing
also we need to find support for infrastructure to run and provide these services

Would like to see the Knowledge Exchange Office to bring together different communities.

Q: who should define service descriptions in a "yellow pages"
A: it's a question of finding a common understanding

Info Grid 2005 - Tuesday 27th, 09:00 - Developing e-infrastructure to support new research and learning paradigms

09:00
Welcome

continuing Service-Oriented theme

Developing e-infrastructure to support new research and learning paradigms
Dr Liz Lyon
Director, UKOLN

Presentation: ppt (5.6M); pdf (5.8M);

focus on research
focus on eBank UK

1. e-Research: a changing landscape

data overload!
how do we disseminate?

diverse types of data

think about how data collections evolve over time

Recommendations from "Large scale data sharing in the life sciences" UK report June 2005
- standards, metadata
- data management
- vocab
- archiving

RCUK open access to data
"should be made available as widely and rapidly as possible"

[diagram of the scholarly knowledge cycle]

2. Developing repositories

Developing models

http://www.e-framework.org/

"service-oriented technical framework... for research and learning"

* reference models
* service definitions

[diagram of JISC Information Environment Architecture - predates e-framework]

[diagram positioning eBank in scholarly knowledge cycle]

eBank UK Project
- open access to datasets
- linking research data to publications and to learning
- JISC funded from Sept 2003: now in phase 2

Exemplar: e-science testbed 'Combechem'
- grid-enabled combinatorial chemisty / crystallography
- national crystallography service

PSIgate (physical sciences info gateway) at Manchester

[diagram of data flow in eBank UK]

[diagram of combechem]

[crystallography workflow diagram]

Crystal Structure Data Reports

http://ecrystals.chem.soton.ac.uk/

data is harvested (OAI), and then aggregated by eBank service

"proof of concept demonstrator project"

*** Linking data to publications ***

done in eBank UK portal

eBank also embedded into PSIgate portal

Issue: Ontologies for discovery in an interdisciplinary world

Issue: Persistent identifiers for data citation

- working on use cases
- various schemes: DOI, handle, ARK, PURL
- there are some identifiers within domains

Publication and Citation of Scientific Primary Data Project

National Library for Science and Technology (TIB)
University of Hanover, Germany

DOI for datasets
http://www.std-doi.de/

can cite data using DOIs

Integration into crystallographic publishing practices

working with IUCr journals

Integration into chemistry research workflows

* R4L - Repository for the Laboratory
* SMART TEA electronic lab notebook + annotations, myTea project

* How does this fit with research assessment (RAE) process?  (UK process)

Integration into the curriculum and e-Learning workflows

* MChem course
* assess role in undergrad chem courses
* introducing school children to e-research?

Knowledge extraction and post-processing

* mining (data, text, structures)
* modelling
* analysis
* presentation (visualisation, rendering)
* in federated repositories: digital libraries, datasets, learning materials
* role of Google??

Repositories and digital curation

data preservation is in conflict with data curation (active use)

http://www.dcc.ac.uk/

upcoming DCC conference September 29-30, Bath, UK

http://www.dcc.ac.uk/training/dcc-2005/

more info
http://www.ukoln.ac.uk/

Q: How are you making the links between the publications and the datasets
A: we're looking for identifiers to identify the datasets to enable linking
who does that? publishers? institutions (within repository)?

currently in the demonstrator, they have a schema,
but they are looking at automated ways

Comment: Germany - subject-oriented information nets - initiative from the researchers themselves
They choose and organize the info they way they want.  Not that well-structured, but they answer a
bit more the needs of the researchers

Info Grid 2005 - Tuesday 27th - Monday summary on evolution of libraries online

I found Carl Lagoze's talk on the next-generation NSDL particularly interesting.  Basically he talked about two stages in the evolution of the understanding of digital libraries.  In the first phase, people were just striving to get content online.  At that time they maybe couldn't even imagine that within a few years, the Internet would have vast amounts of information available.  In some ways, people were proceeding from an imperfect model of the library:

The formal part of a library is books come in, get catalogued, and then people use the card catalogue to locate books on the shelves.

This was replicated in content repositories (digital libraries, phase 1).

But this missed the whole informal, social aspect of the library.  You lose both the community and the context when you just have raw indexed content online.

If I can turn a phrase, in a library, people congregate around other people, not around the card catalogue.

So in phase 2, they are attempting to bring back the contextual and social aspects of the library.  This means that what some people may consider frivolous applications: blogging, photo sharing, instant messaging, recommendations, bookmarks... all these aspects of person-to-person communication need to be integrated.  As well as automated "wisdom of crowds" technology: "the person who read this paper also read these papers" etc.

The good news is, there is no need to reinvent the wheel - the whole point of the Service-Oriented Architectural (SOA) methodology is that your role becomes as much one of integration as of creation.  Take the best of the online services, and integrate them into your site.

The other part of the SOA paradigm is, in a way, that small, simple, imperfect but quickly developed and released pieces of technology can be more successful than complex, perfected tech.  "Good enough" can be great.

I'm actually a bit concerned that we may get too caught up in elaborate Web Services architectures when simpler protocols may in some cases work better.  But we have to balance that against getting trapped in silo solutions.  The key is to architect so that components are loosely coupled.  Herbert Van de Sompel showed how doing a good internal architecture with strongly separated components means that those components can then be re-used in external applications and distributed environments.  In effect, he treats his local network as if it is a distributed network.

InfoGrid 2005 - photos

I am Flickring photos under tag infogrid2005.  Had some major problems with Flickr last night, but things are working ok this morning.

Info Grid 2005 - Tuesday 27th - presentations are up

(UPDATE: Most of the presentations from both days)  at Building the Info Grid 2005 are up, they are on the speakers page, just below the speaker's URL and above the abstract.  PowerPoint and PDF formats.  This is a wonderful effort by the conference organizers, as I have been to other conferences where it took weeks for the presentations to show up on the website.

Info Grid 2005 - Monday 26th, 4:30 - JISC-DFG-SURF-DEFF Knowledge Exchange Initiative

4:30
JISC-DFG-SURF-DEFF Knowledge Exchange Initiative
Four Nation Collaboration Agreement to Support ICT Development in Education and Research
Diann Rusch-Feja
Knowledge Exchange Office, Copenhagen

Denmark (DEFF) - Denmark's Electronic Research Library
Germany (DFG) - German Research Foundation
Great Britain (JISC)
Netherlands (SURF)

Mission, Vision and Goals

Mission
- support innovative devel and use of ICT in edu and research
- increase return on investment
- enhance services and project through greater knowledge dissemination

Goals
- increase profile of national research
- where appropriate, strive for common infrastructure based on common standards

Org
- two delegates (board members) from each country selected by national funding agencies + chair from hosting country (Denmark)
- national representative from each country

will look at additional countries in Europe and outside Europe...

http://knowledge-exchange.info/ (not up yet)

Official Launch December 2005.