Ana səhifə

National Library of Medicine Recommendations on nlm digital Repository Software

Yüklə 1.4 Mb.
ölçüsü1.4 Mb.
1   2   3   4   5   6   7   8   9   ...   15

5. Recommendations

5.1. Recommendation to use Fedora and Conduct a Phase 1 Pilot

The Digital Repository Evaluation and Selection Working Group recommends Fedora as the core system for the NLM digital repository and to start now on a phase 1 pilot to involve real collections. Fedora's architecture should enable NLM to ingest, manage, and deliver exotic content as well as the typical digital scans of print originals. It has the potential to encourage creative approaches to digital library research and development, e-publishing, e-scholarship, and e-science.

Fedora has been implemented by a number of institutions involved in innovative digital services, including Indiana University, Rutgers University, Tufts University, the University of Virginia, the Max Planck Society (eSciDoc), the National Science Foundation (The National Science Digital Library), the Public Library of Science, and the Inter-University Consortium for Political and Social Research.

Drawbacks include the extensive customization, training, and support required to implement and manage the complex architecture. Considerable time also will be invested in developing detailed workflows for Fedora. These risks, while significant, do not outweigh the system's benefits.

5.1.1 Key reasons for Fedora

  • Provides the flexibility that will be needed to handle NLM's near-term and foreseeable future needs.

  • Has a strong technology roadmap that is aggressively advancing scalability, integration, interoperability, and semantic capabilities.

  • Is being used by leading institutions that have digital projects goals similar to NLM's.

  • Has an active open source development community that is well-funded with grant money. Fedora is cutting edge yet bounded by a strong commitment to standards.

  • Strongest and most flexible metadata support of all candidates - it is not bound to any single scheme.

  • Hands-on functional testing has demonstrated that Fedora by itself scored well against NLM functional requirements, and, with the Fez add-on front-end tool, scored higher than DSpace and DigiTool.

  • Fedora is a low-risk choice for NLM at this time:

    • Fedora is open source software, so there are no software license fees.

    • Other institutions like NLM are building effective digital collections using Fedora, and they can provide valuable advice and lessons-learned.

    • Digital assets ingested into Fedora can be easily exported, if NLM were to decide to take a different direction in the future.

    • Fedora is a good opportunity for NLM to gain experience with open source software.

    • Fedora is developed and maintained using technologies that OCCS can support.

5.1.2. Future Actions Needed

After the completion of a pilot, NLM should evaluate its work. Evaluation is a prudent plan to mitigate any risks associated with using Fedora. The pilot group should also re-evaluate the repository software landscape as new versions of all the tools examined are coming out over the next 12 months, including:

  • Fedora just released version 3.1 which makes significant improvements in defining the content model.

  • DSpace architecture will undergo major improvements with a new version, DSpace 2.0.

Plans are also being made for significant collaboration between the DSpace and Fedora communities and NLM should keep abreast of how these plans could support NLM's use of Fedora.

The pilot group may also want to determine if NLM should conduct a formal test of the Ex Libris Digital Preservation System (DPS). DPS is an emerging new commercial tool that offers future promise for digital repository applications:

  • DPS is being developed to meet the requirements of the National Library of New Zealand (NLNZ), which rejected DigiTool.

  • Release 1.0 is expected to be generally available by end of 2008/early 2009.

  • NLNZ has gone live with DPS and is happy with the results so far.

5.2. Phase 1 Pilot Recommendations

NLM should start with Fedora 3.1, the latest production release version. NLM hasn't exhaustively tested 3.x but is starting to examine the code and new key features. Other institutions which the group has spoken with are planning to migrate from 2.x to 3.x.

5.2.1. Companion Tools

  • Use of Fedora open source software gives NLM the opportunity to select and incorporate "best-of-breed" companion tools.

  • NLM can replace or add new tools as better alternatives become available.

  • Tool awareness, evaluation, and selection will be a part of NLM's repository evolution process.

  • Companion tool investigation needed during phase 1 pilot:

    • Administrative interface tools: The pilot group should not commit immediately to Fez but should investigate alternative administrative interface tools such as Muradora or the Rutgers Workflow Management System.

    • Preservation tools: Determine use of JHOVE and related tools such as DROID for file identification, verification and characterization.

    • Public user interface tools: Research and implement either open source or commercial page turning or other front end access capabilities and software.

5.2.2. Workflows

  • The pilot group should make workflow recommendations over time and workflows may be tied to the collection or type of material.

  • Workflows to be initially examined probably include metadata needed for SIPs (Submission Information Package) and format characterization.

5.2.3. Suggested Phase 1 Pilot Scope and Time Frame

6-8 months:

  • Develop a first pilot collection that already has metadata and associated files. Produce a "quick" success to show progress.

  • Manage the content in one secure place.

  • Focus on defining the core functions in the areas of: data models, metadata, preservation and SIP creation.

  • Investigate interfaces with Voyager to maximize use of existing metadata.

  • Provide an initial public presentation using a simple Web interface.

  • Investigate and begin to implement key preservation aspects to ensure master files are preserved.

8-18 months:

  • Implement an additional one or two pilot collections (of the 4 proposed in section 5.4).

  • Begin making recommendations on institutional workflows.

  • Implement an administrative interface or collaborate with other users to evolve some open source alternative, or integrate/develop our own.

  • Implement one or two unique public access capabilities (e.g., a page turning application).

5.2.4. NLM's Role in the Fedora Open Source Community

  • NLM should investigate potential participation in the Fedora Commons community, e.g., the Fedora Preservation and Archiving Solution Community group.  Participation could enable NLM to influence future software features.  NLM should also investigate potential partnerships with leading Fedora users, e.g., University of Maryland, University of Virginia, or others.  (These are strategic/management decisions.)

  • NLM should consider contributing source code to the Fedora community only after the pilot phase, if NLM decides to continue its use of Fedora. NLM should become a participant rather than a "lurker."

  • Before NLM shares any code it may want to consult with NIH legal counsel.
1   2   3   4   5   6   7   8   9   ...   15

Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur © 2016
rəhbərliyinə müraciət