Notes
Slide Show
Outline
1
Interoperability and Reuse
  • David Millman
  • Columbia University
  • dsm@columbia.edu
  • April 2006
2
Projects
  • DART - Anthropology teaching resources
  • Web services modeling
  • NSDL Services
  • Journal of Jazz Studies
3
DART Library

  • Narrative navigation from instructor


  • Faceted navigation from metadata
4
Narrative Navigation
5
Navigation Choices
6
Navigation via metadata
7
Metadata facets
8
Library/Repository Services
  • Library offers consistent metadata for all items used by the instructor
  • Library should offer consistent service interface
  • What are the right services? (e.g., Sakai twin peaks for search-type discovery)
9
Sample “Service-Oriented” Architecture
10
Decisions
  • Repository definition / collection scope
  • Service definition
  • Publication definition
  • Delivery to non-web-browsers (software agents, mobile devices)
  • Publication as services, not web site?
11
NSDL Services Enabled
(National Science Digital Library)
  • Multiple collections == Multiple sites
  • Union catalog interface
  • Columbia role: Single sign-on
  • Implications of seamless login between sites: other seamless services?
12
NSDL Architecture
13
Possible NSDL Architecture
14
Lessons so far
  • Single sign-on suggests more
  • Machine-machine services offer possibilities
  • Narrative/constrained vs free roaming the new stacks and the Internet
  • Define services & exchange formats
15
Personal Library vs.
Research Library (Print)
  • Individual selection criteria
  • Personal shelving system understood only by individual
  • Decision to lend material made by library owner
  • No backup preservation
16
Individual Web-Environments vs. Scalable Digital Library Resources
  • Develop individual digital teaching tool


  • Broad range of approaches to citation and permissions


  • Individual systems for tagging content


  • Organize digital content for individual class


  • Storage and access dependent on individual author



17
Next Steps
  • Journal of Jazz Studies  (Fall ’06, Ford Foundation funding)
  • Investigate service  requirements through workflow
  • “Spectrum of stability”
18
Possible Spectrum
19
Research Questions
  • How many discrete stages?
  • What features are appropriate at each stage?
  • What actions happen as content crosses stage boundaries?
  • Can we create service requirements from these features?
20