The centuries old exchange of culture and commerce along the San Antonio River is an apt metaphor for The Second International Open Repositories Conference in San Antonio held at the River Center Marriott. The conference theme, “Achieving Interoperability in an Open World” echoed objectives of early settlers who came to the riverbanks in the late 1600s looking for homes and prosperity in an untamed land. The 350 attendees at the January 23-26 event found a home in an evolving community anxious to exchange everything from tourist information, collaborative solutions, code and ultimately institutional knowledge.

From a Fedora keynote address by Kaye Howe (see this blog) to presentations about collaborative infrastructure and community development, NSDL emerged as a significant partner in the open source community. Core Integration members Dean Krafft, Aaron Birkland, Chris Wilper, and Carol Minton Morris presented during the Fedora User group sessions. Cathy Lowe and Laura Bartolo gave a poster at a lively poster reception entitled “NSDL MatDL: Adding Context to Bridge Materials E-research and E-education.” NSDL’s distributed role in solving technical and social issues around the development of institutional knowledge was underscored during a plenary presentation by Huda Kahn and Keith Maul from DLESE entitled,”Realizing the Role of Digital Repositories in Educational Applications: Supporting Content and Context.”

All three organizing repository projects—Dspace, Fedora and Eprints—hosted concurrent user group sessions followed by plenary talks focused on overarching issues and solutions. Attendees were encouraged to “cross-pollinate” among user group sessions.

NSDL Presentations

NSDL 2.0: Building a Collaborative Digital Library
DEAN KRAFFT gave the kick-off presentation at the Fedora User Group sessions with a look at the NSDL 2.0 Fedora-based NSDL Data Repository and emerging early implementations of collaborative tools.

He began with a brief overview of NSDL and a summary of requirements for creating the NSDL Data Repository based on Fedora:
-Implemented in Fedora 2.1+ with MPTStore and journalling
-References to over 2 million selected STEM resources on the web, plus some content local to the repository
-Sourced metadata statements about those resources
-A REST API to allow authenticated access by Pathways, providers, tool builders
-In production at NSDL.org

He reviewed an NSDL view of “collaboration, context, and contribution” emphasizing successful models—blogs, wikis, bookmarking/tagging; successful software—Wordpress, Media Wiki, Connotea and; successful community practices—engaging with partners and the broader community to build applications on the NDR platform (adapting DLESE collection system, integrating Instructional Architect)

He demonstrated Expert Voices entries utilizing plug-ins developed by Elly Cramer to link references to NSDL resources and add references to URLs that should become resources.

In conclusion Krafft highlighted other collaborative tools in development including “Our NSDL”, “My NSDL” and the “challenges in creating a collaborative repository of resources, metadata and context” while showing an old photograph of a barn raising. The structure of the barn is visible behind a large crowd of community builders where successful completion clearly depended on trust and reputation management, as well as hard work and cooperation. These qualities are also needed to create a collaborative web of context in NSDL 2.0.

“It’s All in How You Slice It,” Fedora Outreach and Communications: A Working Group Report
CAROL MINTON MORRIS AND STACY PENNINGTON At the Fedora Users Group Meeting held at the University of Virginia in June 2006 two central outreach and communications questions emerged for the group to tackle, “Who is the Fedora Community of projects and can we engage it further to increase adoption of the core technology?” From September 2006-November the Fedora Outreach and Communications Working Group conducted a study to begin to answer those questions and find out what distributed Fedora deployment and development means in the context of a wide variety of use cases. Presenters reviewed how the study was conducted, highlighted results and showed the community back to itself with an analysis of what an aggregated Fedora community looks like.

The study had a 30% response rate. Participants were contacted by phone and through email. In response tone question, “Why did you chose Fedora” respondents replied (in ranked order):

1. Flexible, extensible, open
2. An architecture to build on
3. Institutional and consortium need to work with open source software
4. Peer development opportunities
5. Service based architecture
6. Technical evaluation of appropriateness for research repositories

A “project sort” revealed that most of the known Fedora community was comprised of large (foreign national, consortium and statewide projects) and small (individual projects or colleges) institutional projects collaborating to extend repository access along four axes:

Overall, members of the current Fedora community were found to be somewhat homogeneous with respect to their roles within their organizations. Respondents were primarily lead technical developers who were responsible for implementation strategies including analysis and planning.

Although Open Source development was important, Fedora was primarily selected because the architecture supported institutional goals over the short and long-term. “Fedora seemed ‘future proof’–able to accommodate whatever technologies would come along.”

Most respondents saw themselves as active members of the Fedora community. Some respondents also saw a need to establish communities of practice focused on rapid development of solutions to micro community problems.

MPT Store: A Fast, Scalable, and Stable Resource Index
CHRIS WILPER AND AARON BIRKLAND co-developed a fast, scalable RDMBS triplestore for the NSDL Data Repository that was also released as part of the Fedora 2.2 software. This work was motivated by the NSDL use case:

1. NSDL has a moderately large repository of 4.7 million objects and 250 million triples and has a large volume of writes
2. Driven by periodic OAI harvests
3. Primarily mixed ingests and datastream mods
4. Highly concurrent reads and writes

The RDMBS triplestore supports much faster simple queries and a similar performance for complex queries.

Realizing the Role of Digital Repositories in Educational Applications: Supporting Content and Context
HUDA KAHN AND KEITH MAUL presented DLESE Teaching Boxes–customizable, digital replicas of the traditional collections that most educators create, store (in boxes), re-use and improve on during their years of teaching. Khan and Maul from DLESE: Digital Library for Earth System Education, talked about bridging the gap between rich digital library resources and usable materials for in-classroom teaching with Teaching Boxes.

Kahn explained that this service is analogous to the cardboard boxes that teachers collect lessons in, complete with materials and instructions. The 2006 version of this must somehow link to student background information, student literacy standards, take into account time contraints as well as teacher backgrounds. Teaching boxes incorporate content and context (photograph of Sandy Payette and Huda Kahn).

Teachers came up with this concept. Interviews with teachers showed that re-use was just one aspect of using these types of teaching materials—adaptation, a kind of tweaking of the materials to meet a teacher’s particular needs, was more likely.

Maul reviewed development of the Teaching Box Builder architecture and application. They needed a model that would adjust according to needs of teachers so flexibility was particularly important. He went on to explain that search capabilities will be a significant aspect of any release version of this software.

They concluded with a live demonstration of an early Teaching Box implementation and how application design provided support for educators’ creation and adaptation of pedagogical content and context.

Near the end of the even John Legget, OR07 Conference Co-chair, said, “If only lip service is given to the conference theme, ‘Achieving Interoperability in an Open World,’ that would be a threat to the community we have seen emerge here.” The work presented by the NSDL community at ORO7 is contributing to making a dynamic web of STEM content and context interoperable, enabled by the release of NSDL 2.0.

Additional Notable Links

1. Presentations from OR07
(available on the web site in the next week or so):
http://openrepositories.org

2. Disruptive Library Technology Jester (We’re disrupted, We’re Librarians, and We’re Not Going to Take it Anymore) summaries:

Wilper, Birkland, MPT Store: A Fast, Scalable, and Stable Resource Index
http://dltj.org/2007/01/fedora-mptstore/

Rogers, Smith, Cross-Repository Semantic Interoperability: the MIT SIMILE Project
http://dltj.org/2007/01/simile/

Payette, A Vision for Fedora’s Future, An Implementation Plan to Get There, and A Project Update
http://dltj.org/2007/01/fedora-update/

3. Howe, Risk, What Risk: Choosing Fedora for the National Science Digital Library
http://expertvoices.nsdl.org/roadreports/

4. A conference-goer’s page for tagging relevant OR07 content:
http://del.icio.us/tag/or2007

5. Dspace Federation
http://www.dspace.org/

6. Fedora Project
http://fedora.info

7. Eprints
http://www.eprints.org/

8. Third Annual Open Repositories Conference 2008
http://openrepositories.org/2008