JSTOR Advanced Technology Research Group

1
JSTOR Advanced Technology Research Group

• Working in collaboration with other researchers
  to provide access to advanced technologies within
  the same workspace as the literature and primary
  source material.
• This will enhance discovery and further
  encourage the use of the materials, enabling new
  scholarship and the creation of new knowledge.

2
JSTOR

• As a digital library weve done quite well.

3
(No Transcript)
4
(No Transcript)
5
Research Support

• JSTOR has always worked with researchers by
  providing datasets and supported them in
  analyzing our data in their research.
• We continue to that and will rarely refuse a
  reasonable request (we handle1-2 dozen requests a
  year for usage, citation and content data).
• Recently we decided to become a more active
  participant in the use of JSTOR data for research.

6
Evolving JSTORs Technology

• JSTOR is, for many scholars, their digital
  bookshelf (or part of it).
• The real work takes place at the workbench, not
  the bookshelf
• Workbenches are trade specific (though tools
  need not be).
• We need to recognize the diversity of practice
  and, as yet, neither we nor the practitioners
  really understand what is needed for each digital
  practice.
• JSTORs Showcase is where we bring technology and
  scholarship together and develop digital
  workbenches for our constituents.
• It is an open, ongoing digital workshop with
  shared tools and materials, where we try to build
  workbenches with and for other scholars.
• It will be extended to interwork with other
  facilities and provide APIs to our functionality
  and resources.

7
The plan

• We will host tools and technologies from the
  community (including JSTOR), quickly and openly,
  working on JSTOR content (and others where
  available).
• Showcase will be a step toward real offerings,
  but our betas will be as useful and usable as we
  can manage.
• We will actively solicit and respond to feedback,
  so that our workbenches will evolve.
• We will provide a place where researchers can
  expose their work to users for the mutual benefit
  of both.

8
Active projects

• DfR Simple text mining and corpus exploration
• Visualizations of JSTOR usage, participants
• Topic mapping ( Blei / Princeton )
• Document Remastering from camera Images aka
  Decapod (Breuel/Kaiserslautern,
  Treviranus/Toronto)
• Open Annotation Collaboration (Cole, Von de
  Sompel, Cohen, Sanderson et al)

9
Data for Research - Examples

• The long s
• The British Empire
• The golden age of social sciences

10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
Foresite

• A collaborative program with University of
  Liverpool and HP Labs, Bristol
• Build a relationship graph of the entire JSTOR
  corpus.
• Explore using an acetate overlay model.

14
Foresite OAI-ORE Explorer. With U. Liverpool
HP Labs
15
Decapod

• A collaborative program with University of
  Kaiserslautern University of Toronto. Funded by
  the Andrew Mellon Foundation.
• Building a small, inexpensive, easy to operate,
  1-click paper to document digitization rig.
• Apply state of the art document understanding and
  usability to allow small institutions to digitize
  their collections.

16
Decapod

• 1-Click, paper to remastered document.
• OSS software.
• state-of-art document understanding.
• Mobile friendly (reflow).
• Operator friendly.
• Budget Friendly.
• Based off Ocropus Fluid
• Partners DFKI/Kaiserslautern, ATRC/Toronto,
  JSTOR

17
Open Annotation Collaboration

• A Mellon funded project, starting in May 2009,
  with the over arching goals to
• Facilitate the emergence of a Web and
  Resource-centric interoperable annotation
  environment that allows leveraging annotations
  across the boundaries of annotation clients,
  annotation servers, and content collections.
  Interoperability specifications will be devised.
• To demonstrate through implementations an
  interoperable annotation environment enabled by
  the interoperability specifications in settings
  characterized by a variety of annotation
  client/server environments, content collections,
  and scholarly use cases.
• To seed widespread adoption by deploying robust,
  production-quality applications conformant with
  the interoperable annotation environment in
  ubiquitous and specialized services and tools
  used by scholars (eg. JSTOR, Zotero, and MONK).
• The partners in this project are
• University Library and Graduate School of Library
  and Information Science, University of Illinois
  at Urbana-Champaign
• Center for History and New Media, George Mason
  University
• Maryland Institute for Technology in the
  Humanities, University of Maryland
• eResearch Laboratory, School of Information
  Technology Electrical Engineering, The
  University of Queensland
• Research Library, Los Alamos National Laboratory
• JSTOR
• Contacts
• Tim Cole UIUC
• Clare Llewellyn JSTOR

18
What we are investigating.

• Corpus Analysts Workbench
• Topic Mapping with LDA (Blei / Princeton)
• Extract topic signatures and trace them through
  time.
• Evolution of Ideas (Blei Gerrish / Princeton)
• Identify ideas.
• Concept extraction using Associative Rule Mining
  (Sanderson / Liverpool)
• Citation strategies and the impact of those
  strategies (Adamic / UM)
• Citation and Similarity structures in Corpora
  (Bergstrom West / UW )
• Eigenfactor and similar techniques.

19
What we are investigating (cont.. )

• Linguists Workbench
• Tools for Linguists using SEASR (Llora / UIUC,
  ODonnell / U. Mich.)
• Association Rule Mining from the journal
  Evolution (Sanderson / Liverpool).
• Oldest English Words analysis (Pagel / Reading)

20
What we are investigating

• Computer-Assisted analysis and mineable Knowledge
  bases
• Art Auction Catalogues
• Using accredited crowd-sourcing to review,
  correct machine-generated corpus of documents
  and lot records.
• Computer Aided Transcription of manuscripts.
• Using modern image-matching techniques to
  scattershot transcribe documents _and_ build a
  database of script signatures for mining.
• Digital Staining of Plant Specimens.
• Use feature extraction, analysis and
  false-coloring to highlight morphological
  structures and create a database of signatures
  for mining.

21
Base Image
22
Analysis from previous illustration showing
reticulated venation.
23
What we are investigating

• Community Specific / Miscellaneous
• Tools for Secondary Schools (SEASR) (Llora /
  UIUC)
• Reading level synopsis generation.
• NEH/NSF/SSHRC/JISC Digging into Data program
  (community support participant in 1 proposal)

24
Summary

• Thank you
• John Burns john.burns_at_jstor.org
• Main site http//www.jstor.org
• Showcase http//showcase.jstor.org
• Data for Research http//DfR.jstor.org