1. The ARROW Project: A consortial institutional repository solution, combining Open Source and proprietary software David Groenewegen ARROW Project Manager

2. DORDSL Workshop, 21 September 2006 2 Outline  Why did we want a repository?  What is ARROW?  What is VITAL and how does it relate to Fedora?  Where is ARROW now?  What have we learnt so far?  ARROW Stage-2

3. DORDSL Workshop, 21 September 2006 3 Why did we want a repository?  provides a platform for promoting research output in the ARROW context  safeguards digital information  gathers an institution’s research output into one place  provides consistent ways of finding similar objects  allows information to be preserved over the long term  allows information from many repositories to be gathered and searched in one step  enables resources to be shared, while respecting access constraints (when software allows access controls)  enables effective communication and collaboration between researchers

4. DORDSL Workshop, 21 September 2006 4 What is ARROW? ARROW Project:  Originally funded for 3 years until December 31, 2006, recently extended for 12 months.  Funded by the Australian Commonwealth Department of Education, Science and Training (DEST), under the Research Information Infrastructure Framework for Australian Higher Education. “The ARROW project will identify and test software or solutions to support best practice institutional digital repositories comprising e-prints, digital theses and electronic publishing.”

5. DORDSL Workshop, 21 September 2006 5 Who is ARROW? Founding ARROW Partners:  Monash University (lead institution)  National Library of Australia  The University of New South Wales  Swinburne University of Technology. ARROW Members:  University of South Australia  University of Southern Queensland  Queensland University of Technology  Central Queensland University  University of Western Sydney  La Trobe University  4 other RUBRIC members are expected to sign soon Together they form the ARROW Community

6. DORDSL Workshop, 21 September 2006 6 What did the ARROW project set out to achieve?  Solution for storing any digital output  Initial focus on print equivalents – theses, journal articles  Now looking at other datasets, learning objects  More than just Open Access – some things need to be restricted  Copyright  Confidentiality/ethical considerations  Work in progress

7. DORDSL Workshop, 21 September 2006 7 What did the ARROW project set out to achieve? (2)  Meeting DEST reporting requirements  Expected move to Research Quality Framework (RQF) has increased the focus on repositories  Employ Open Standards  Making sure the data is transferable in the future  Deliver Open Source Tools back to the FEDORA Community  Solution that could offer on-going technical support and development past the end of the funding period

8. DORDSL Workshop, 21 September 2006 8 What is ARROW now?  A development project  Combining Open Source and proprietary software:  Fedora™  VITAL  Open Journal Services (OJS)  NOT a centralised or hosting solution  Every member has their own hardware and software

9. DORDSL Workshop, 21 September 2006 9 Why Fedora? ARROW wanted:  a robust, well architected underlying platform  a flexible object-oriented data model  to be able to have persistent identifiers down to the level of individual datastreams, accommodating its compound content model  to be able to version both content and disseminators (think of software behaviours for content)  clean and open exposure of APIs with well-documented SOAP/REST web services.

10. DORDSL Workshop, 21 September 2006 10 ARROW and Fedora™ Since the beginning of the project ARROW has worked actively and closely with Fedora™ and the Fedora Community  ARROW Project Technical Architect is a member of Fedora Advisory Board  ARROW Project Technical Architect sits on Fedora Development Group This is reinforced by VTLS Inc.  VTLS President is a member of Fedora Advisory Board  VITAL Lead Developer sits on Fedora Development Group

11. DORDSL Workshop, 21 September 2006 11 Partnering for success, support and survivability  ARROW needed to partner with a developer who could not only produce the software but could provide ongoing user support and development after December 31, 2006  Why VTLS Inc.?  VTLS wanted to be a development partner  Had begun work on a repository solution already  Familiar with library sector  Willing to produce a combination of a proprietary solution, Fedora and other Open Source software

12. DORDSL Workshop, 21 September 2006 12 What is VITAL? ARROW specified software created and fully supported by VTLS Inc. built on top of Fedora™ that currently provides:  VITAL Manager  VITAL Portal  VITAL Access Portal  VALET - Web Self-Submission Tool  Batch Loader Tool  Handles Server (CNRI)  Google Indexing and Exposure  SRU / SRW Support

13. DORDSL Workshop, 21 September 2006 13 Fedora™ VITAL architecture overview Indexes Handles server Web services Google exposure SRU/SRW Batch Loading Tool Access Portal Valet Vital Manager

14. DORDSL Workshop, 21 September 2006 14 Where are we now? 2004 Developed architecture Selected, tested Fedora™ OJS VITAL 1.0 2005 VITAL 1.3 Started populating repositories OAI-PMH harvesting ARROW Discovery Service Open sources tools released VITAL 2.0 2006 VITAL 2.1 VITAL 3.0 (in test) Authentication/Authorization Services Enhanced Content Models Usage and access statistics User configurable interfaces Movement towards a pure Web based interface Support for OAI sets Integration with 3rd party modules like federated search 2007 ARROW Stage-2

15. DORDSL Workshop, 21 September 2006 15 ARROW Repositories  Monash University  http://arrowprod.lib.monash.edu.au:8000/access  University of New South Wales  http://arrow.unsw.edu.au/  Swinburne University of Technology  http://researchbank.swinburne.edu.au/access/  Central Queensland University  http://library-resources.cqu.edu.au:8888/access/

16. DORDSL Workshop, 21 September 2006 16 Implementation decisions  Atomistic or compound objects?  Descriptive metadata  adopt one or enjoy MANY types?  JHOVE validation  JHOVE metadata extraction  Use cases and content modelling  What import /export formats?  honouring what standards?  validation, when and how?

17. DORDSL Workshop, 21 September 2006 17 Policy frameworks and decisions  Direct or mediated deposit?  managing workflows  Open or closed access?  LDAP authentication?  XACML authorisation  creating policies -who can do what?  Shibboleth  Persistent URL format?  External searching and harvesting?  OAI-PMH  spidering  post ARROW project support  For more detail see Andrew Treloar’s talk at: http://www.lib.virginia.edu/digital/fedoraconf/schedule.shtml http://www.lib.virginia.edu/digital/fedoraconf/schedule.shtml

18. DORDSL Workshop, 21 September 2006 18 External searching and harvesting  Realised need to develop a discovery service for Australian institutional repositories  The ARROW Discovery Service developed by the NLA, provides consolidated searching across many Australian repositories, (uses OAI-PMH)ARROW Discovery Service  Picture Australia developed by the NLA, harvesting image collections (uses AOI-PMH) Picture Australia  SRU/SRW interface released as Open Source Software  Harvesting  Google and other service providers

19. DORDSL Workshop, 21 September 2006 19 ARROW Discovery Service http://search.arrow.edu.au/  Provides a national resource discovery service including:  providing an appropriate search interface  simple search, advanced search, & browse options  contributing to other networks  OAIster, Yahoo, Google  Ensuring appropriate local institutional and national “branding” of the service  occurs throughout the ADS interface and the exchanged metadata  providing appropriate subject-based access  The Australian Standard Research Classification list

20. DORDSL Workshop, 21 September 2006 20 Open Source contributions for Fedora  Already made:  SRU/SRW  HANDLES  JHOVE Metadata extraction  Exposure to Web indexing crawlers.  Coming in 2006:  LDAP Authentication  Administrative Reporting  Bulk Citation Export  Statistics for Public Users  Metadata Synchronisation Requirements

21. DORDSL Workshop, 21 September 2006 21  Upcoming VITAL Version 3.0  Authentication/Authorization Services.  XACML (Policy enforcement)  Enhanced Content Models.  Usage and access statistics.  User configurable interfaces.  Movement towards a pure Web based interface.  Support for OAI sets.  Integration with 3rd party modules like federated search.  Access to content via VTLS reseller arrangements. Future of VITAL

22. DORDSL Workshop, 21 September 2006 22 What have we learnt so far?  Multiple partners are good:  Sharing of information and experiences  Sharing of development work  Multiple perspectives on issues  and bad:  Multiple perspectives on issues  Scope creep  Managing expectations  Pressure on the project management team  Pressure on development team and partners  Deadline conflicts  Software development feels slow, both commercial and open source  Development with a commercial partner can be tricky

23. DORDSL Workshop, 21 September 2006 23 What have we learnt so far? (2)  That there aren’t enough real standards in this area  Open versus closed repositories, or information management versus accessibility is a BIG ISSUE  Repositories are only partly about software - advocacy, policy, institutional engagement and grunt work need equal attention  Constraints of dealing with copyright

24. DORDSL Workshop, 21 September 2006 24 ARROW Stage 2  Funded to the end of 2007  Supporting the RQF  Creative development of institutional repositories  Supporting Australian engagement with institutional repositories  Building partnerships to further enhance repositories  Identifier Management Infrastructure for e-Research Resources (PILIN)

25. DORDSL Workshop, 21 September 2006 25 Some changes in direction  Trying to do more development ourselves to:  Spread the knowledge  Leverage our use of Fedora  Want to work with VTLS in new ways  Contract is finished now  Some work we need to do is too local for VTLS  VINES

26. DORDSL Workshop, 21 September 2006 26 Supporting the RQF  Inclusion of all discrete pieces of evidence, regardless of content type  Including traditional text evidence and less traditional evidence, such as art works and music compositions or performances  Provision for maximum possible exposure of content  Subject to copyright constraints.  Inclusion of metadata and links to content in commercial resources.  Reporting to DEST through multiple channels  Such as Research Master, or direct to the repository.  Support for access and authorisation regimes.  Retention of all evidence  To build institutional research profiles over time.

27. DORDSL Workshop, 21 September 2006 27 Creative development of ARROW institutional repositories  Inclusion of multimedia and creative works produced in Australian universities.  To date have had limited exposure nationally or internationally.  Addition of annotation capability  Inclusion of datasets and other research output not easily provided in any other publishing channel.  In conjunction with the DART (ARCHER) Project.  Exploration of the research-teaching nexus by facilitating multiple uses of content held in repositories.  Integration with or development of new tools that will allow value added services for repositories.  For instance the creation of e-portfolios or CVs of research output of individual academics.

28. DORDSL Workshop, 21 September 2006 28 ARROW Projects  ARROW is planning a number of local projects targeting local and community needs. These will interact directly with Fedora™ and VITAL where appropriate. The development is being done within the ARROW Community.

29. DORDSL Workshop, 21 September 2006 29 Partner projects in 2007  Gathering research output from websites (UNSW)  Displaying outputs through websites (portfolios) (UNSW/Swinburne)  Understanding workflows and needs of academics (UNSW/Swinburne)  Improving the ARROW Discovery Service (NLA)  OAI Sets support  Greater automation  Statistics capture and reporting  Integration of e-journals  Usability analysis (Swinburne)  Data needs survey (Swinburne)  Building Rules for Access to Controlled Electronic Resources (BRACER) (Monash)

30. DORDSL Workshop, 21 September 2006 30 Supporting Australian engagement with institutional repositories  FRODO and MERRI projects have resulted in a significant leap in the levels of understanding and engagement with repositories in Australia,  Now the challenge is to translate this into substantial repository activity.  The newly formed ARROW Community is intended to provide a central platform for support and the exchange of information.

31. DORDSL Workshop, 21 September 2006 31 The ARROW community  Sharing knowledge and experiences  Annual meeting – inaugural one September 8, 2006  Regular workshops  Working Groups  ARROW Repository Managers Group  ARROW Development Group  Possibly groups for:  Portfolio design  Metadata: METS, MODS, DC and the future  Discussion group  GoogleGroup  ARROW provides logistical and admin support

32. DORDSL Workshop, 21 September 2006 32 Building partnerships to further enhance repositories Through partnerships with other projects ARROW will endeavor to use best practice and new innovations to further enhance Australian repositories beyond their current limitations. These include:  APSR: http://www.apsr.edu.au/http://www.apsr.edu.au/  DART/ARCHER: http://www.dart.edu.au/http://www.dart.edu.au/  ICE: http://ice.usq.edu.au/http://ice.usq.edu.au/  MAMS: http://www.melcoe.mq.edu.au/projects/MAMS/http://www.melcoe.mq.edu.au/projects/MAMS/  OAK-Law: http://www.oaklaw.qut.edu.au/http://www.oaklaw.qut.edu.au/  RUBRIC: http://www.rubric.edu.au/http://www.rubric.edu.au/

33. DORDSL Workshop, 21 September 2006 33 PILIN - Persistent Identifiers and Linking INfrastructure  Growing realisation that sustainable identifier infrastructure is required to deal with the vast amount of digital assets being produced and stored within universities.  This is a particular challenge for e-research communities where massive amounts of data are being generated without any means of managing this data over any length of time.  The broad objectives are to:  Support adoption and use of persistent identifiers and shared persistent identifier management services by the project stakeholders.  Plan for a sustainable, shared identifier management infrastructure that enables persistence of identifiers and associated services over archival lengths of time.

34. DORDSL Workshop, 21 September 2006 34 Questions?  ARROW Project  arrow@arrow.edu.au arrow@arrow.edu.au  http://arrow.edu.au/ http://arrow.edu.au/  ARROW Project Manager  David.Groenewegen@lib.monash.edu.au David.Groenewegen@lib.monash.edu.au