Building the Interspace:
Digital Library Infrastructure for a University Engineering Community

Bruce R. Schatz, PI
University of Illinois
thabing@uiuc.edu

AAAS-95 Annual Conference

February 19, 1995 Atlanta

Goals

•  
• Construct large-scale digital library testbed SGML documents: engineering journals
• Perform research needed to scale this infrastructure (technology/sociology)
• production library (NII model)
• Net interaction (Interspace)
• infrastructure & architecture

Organizations

•  
• University of Illinois at Urbana-Champaign
• Applications (Testbed Development) Grainger Engineering Library Information Center (part of University Library, UL)
• Systems (Infrastructure Development) NCSA Mosaic (part of National Center for Supercomputing Applications, NCSA)
• Evaluation (Sociology Research) Graduate School of Library & Information Science (GSLIS) + Sociology, Economics
• Interaction (Technology Research) GSLIS + Computer Science (CS), NCSA

Principal Investigators

•  
• Bill Mischo, UL, Testbed Lead
• Tim Cole, UL, testbed collection
• Joseph Hardin, NCSA, Infrastructure Lead
• Larry Jackson, NCSA, infrastructure software
• Ann Bishop, GSLIS, Testbed Evaluation
• Bruce Schatz, GSLIS, Technology Research
• executive committee: UL, NCSA, GSLIS, CS

Testbed Collection

•  
• journal/magazine articles in engineering/science
• text/figures/equations/sidebars
• IEEE Computer Society
• APS (American Physical Society)
• AIAA (Aerospace engineering)
• John Wiley & Sons (commercial)
•  
• engineering bibliographic databases
• scientific databases eg NASA CAN

Testbed Users

•  
• faculty/students at Midwestern universities
• professional engineers/scientists
•  
• University of Illinois
• CIC (Big Ten + 1) engineering libraries
•  
• 100,000 users of 100,000 documents

Testbed Software

•  
• Front end Display
  • NCSA Mosaic: User Interface, Net fetch
  • SGML viewer from SoftQuad (Panorama)
  •  
• Back end Search
  • stateful gateway for search history
  • full-text engine from Dataware (BRS)
  •  
• modify standard DTDs for search functions

Testbed Evaluation

•  
• Interviews (Qualitative Data)
  • 100 in Focus Groups
• Surveys (Quantitative Data)
  • 1000 in Years 2 and 4
• Instrumentation (Transaction Logs)
  • 100,000 record structured access
•  
• different methodologies in granularity & scale
• towards nethnography: quanti-> qualitative

Partners

•  
• society publishers:
  • IEEE Computer Society
  • American Physical Society (APS)
  • Amer Inst Aeronautics& Astronautics (AIAA)
  • American Society of Civil Engineers (ASCE)
• commercial publishers:
  • John Wiley & Sons, U.S. News, Tribune Co.
• testbed:
  • SoftQuad, Dataware, Hewlet-Packard
• infrastructure:
  • Spyglass, CNIDR, Microsoft
• evaluation:
  • CNRI, CMU
• technology:
  • U Arizona, U Wisconsin

Technology Research

•  
• towards the Interspace: Net interaction
•  
• Library Science subject classification
• Information Science semantic retrieval
• Computer Science system performance
• Information Systems new architectures
•  
• design/implement next generation system

Library Science

•  
• evaluate subject classifications for Net search
•  
• manual generation by professional indexers
• LC (Library of Congress classification)
• DDC (Dewey Decimal Classification)
•  
• use subject taxonomy as alternative entry
• compare to journal indexes (Engineering Index, INSPEC thesarus) and full-text search

Information Science

•  
• semantic retrieval through concept spaces
•  
• automatic generation of “thesaurus” from text
• co-occurrence matrix for “related” terms
• suggest vocabulary terms across domains
•  
• promising results in molecular biology (WCS)
• will try in broader domain of engineering
• potential to scale to NII and the Net

Information Systems

•  
• next generation digital library NII model
•  
• support Net interaction and analysis
• design/implement Interspace environment
• generic Worm Community System (WCS)
•  
• incorporate research and testbed efforts
• apply to DLI, CAN (NASA datasets), WCS

Interspace Protocols

•  
• structured objects form information space
• linked information spaces form Interspace
•  
• interaction interoperability (type coercion)
• analysis navigation (session recording)
•  
• communication for people interaction (notify)
• wrappers for program interaction (invoke)

Building the Interspace

•  
• merging community information spaces
•  
• molecular biology
• coli, yeast, worms, flies, mice, men
• neuro biology
• moths, molluscs, rats, cats, monkey, man
•  
• science
• biology, physics, astronomy ...
• engineering
• software, computer, aerospace, mechanical, civil, agri...

Go back to the Home Page