Designing for the Discipline: Open Libraries and Scholarly Communication

1
Designing for the Discipline Open Libraries and
Scholarly Communication

• Thomas Krichel
• 2005-05-20

2
about this talk

• Three parts
• normative theory
• RePEc history
• rclis future ideas
• And a final plea all of this needs help.

3
scholarly communication

• is mainly about scholars communicating
• between themselves
• to students, occasionally
• Thus it is essentially a community activity
• Traditionally, there have been two intermediaries
  acting as external agents.
• libraries
• publishers

4
when tradition ends

• Two external shock
• There comes the Internet and reduces distribution
  costs to zero
• There comes computer technology and reduces
  storage costs somewhat
• opportunity sets of community members and
  external agents increases
• Proposition the future depends much on what the
  community members decide. External agents have
  little impact.

5
discipline communities

• Scholars of various disciplines have varying
  habits of research, publication, and evaluation
• It is likely that the Internet will emphasize
  those differences rather than reducing them.

6
examples disciplines with established informal
publishing

• Preprint communities
• Physics ? arxiv.org
• Mathematics ? arxiv.org, partially
• Working paper communities
• Computer Science ? CiteSeer
• (working paper disappearing)
• Economics ? RePEc

7
change is tough

• Change has to come inside the discipline.
• There has to come a pioneering individual who
• is technically well versed
• is managerially smart
• has extraordinary forward thinking
• is willing to take considerable risk with her
  career
• Ginsparg, Krichel, Giles Lawrence are rare

8
and what about libraries?

• Libraries do it systematically wrong
• concentrate on access
• concentrate on readers
• concentrate on documents
• They need to
• move from access to impact
• move from the reader to the writer
• move from documents to people

9
RePEc

• RePEc is a freely available digital library
  related to Economics.
• It does provide for a partial evaluative
  database.
• It is entirely run by a virtual organization of
  volunteer.
• I am the person who got it starting in 1993.
• I skip over history.

10
RePEc principle

• Many archives
• archives offer metadata about digital objects
  (mainly working papers)
• One database
• The data from all archives forms one single
  logical database despite the fact that it is held
  on different servers.
• Many services
• users can access the data through many
  interfaces.
• providers of archives offer their data to all
  interfaces at the same time. This provides for an
  optimal distribution.

11
RePEc is based on 460 archives

• WoPEc
• EconWPA
• DEGREE
• S-WoPEc
• NBER
• CEPR

• US Fed in Print
• IMF
• OECD
• MIT
• University of Surrey
• CO PAH

12
to form a 312k item dataset

• 153,000 working papers
• 157,000 journal articles
• 1,700 software components
• 1000 book and chapter listings
• and the really important stuff
• 7,000 author contact and publication
  listings
• 8,700 institutional contact listings

13
RePEc is used in many services

• EconPapers
• NEP New Economics Papers
• Inomics
• RePEc author service
• Z39.50 service by the DEGREE partners

• IDEAS
• RuPEc
• EDIRC
• LogEc
• CitEc

14
institutional registration

• This works through a system called EDIRC.
• Christian Zimmermann started it as a list of
  departments that have a web site.
• I persuaded him that his data would be more
  widely used if integrated into the RePEc
  database.
• Now he is a crucial RePEc leader.

15
author registration

• It started when funding allowed us to hire a
  student programmer to write an author
  registration system.
• The system went online as "HoPEc" in late 2000.
• It has been renamed "RePEc author service" (RAS)
• In 2002 grant from OSI allows for a rewrite and
  expansion.

16
RePEc author service

• RePEc document data has author names as strings.
• The authors register with RAS to list contact
  details and identify the papers they wrote.
• This is classic access control, but done by the
  authors.
• Currently one in three items in RePEc has at
  least one identified author.

17
LogEc

• It is a service by Sune Karlsson that tracks
  usage of items in the RePEc database
• abstract views
• downloads
• There is mail that is sent by Christian
  Zimmermann to
• archive maintainers
• RAS registrants
• that contains a monthly usage summary.

18
authors' incentives

• Authors perceive the registration as a way to
  achieve common advertising for their papers.
• Author records are used to aggregate usage logs
  across RePEc user services for all papers of an
  author.
• Stimulates a "I am bigger than you are"
  mentality. Size matters!

19
summary keys to success

• Have a small group of volunteers.
• Disseminate as widely as possible.
• Collect precise usage logs.
• Demonstrate to authors and institutions that it
  works for them.
• institutional registration
• author registration

20
rclis

• rclis stands for Research in Computing and
  Library and Information Science.
• It is pronounced as reckless.
• It is a RePEc clone.
• My attempt to show that the same ideas that
  propel RePEc also can work in that area.

21
technical innovation

• RePEc is built on attribute value templates.
• rclis is built on a purpose built format called
  the Academic Metadata Format.
• I set up this format. It is tailor-made to suit
  the needs of rclis and RePEc.
• There is some usage of AMF in RePEc
• RePEc OAI interface
• ernad, the software feeding NEP

22
E-LIS

• It is the largest LIS eprint archive on this
  planet.
• It lives at http//eprints.rclis.org.
• It contains over 2400 documents.
• It runs in Italy but uses a system of national
  editors to feed in material.
• I am one of the US editors.

23
DoIS

• DoIS is a service based on a Spanish LIS
  bibliography.
• It used to run at Manchester computing but moved
  to http//wotan.liu.edu/dois when, because of
  JISC regulations, we had to move from there.
• It contains 13k records, 9k with free full text,
  but the data has many errors.

24
using already existing resources

• There is already a very large computer science
  bibliography called DBLP, see http//dblp.uni-trie
  r.de
• The data has no abstracts. It has some full-text
  links, mainly to toll-gated sites.
• I have done work to convert parts of it to AMF.
• I am now searching if free full text versions of
  the papers exist anywhere on the Web. This is the
  Konz project.

25
the Konz project

• Current state
• I use Google API to search of titles.
• I examine responses and download pages.
• I scan the pages for PDF and Word files.
• I examine the text in the file to find the title.
• Limitations
• pdf and word full text
• conference paper data still being processed
• significant hardware and disk problems.

26
DoCIS

• Konz currently finds 25k papers with free
  versions out of the paper out of a 98k searched.
  Not particularly exciting.
• This data is integrated with DBLP AMF data and
  the result forms a new service called DoCIS.
• DoCIS lives at
• http//wotan.liu.edu/docis

27
DoCIS service

• DoCIS is implemented in mod_perl with swish and
  therefore very fast.
• The web pages are written by XSLT scripts
  directly from the AMF data.
• The service is available to copy from the web, I
  am more than happy to run it on other sites.
• But the most interesting thing are the service
  principles.

28
construction transparency

• DoCIS is an open digital library service because
  it allows users to inspect exactly how the
  service runs
• DoCIS is built using open source software.
• There is a special interface http//wotan.liu.edu/
  strip/docis/ that allows to see almost all
  internal file. Non visible files are specially
  documented.
• The hope is that it may be used for teaching
  purposes.

29
transportability

• Everything in DoCIS is built is such a way that
  it should be easy to move the service somewhere
  else and establish copies.
• The ideas may not make a lot of technical sense
  but it should increase to non-proprietary nature
  of the system.
• Note that this has not been tested --)

30
usage transparency

• All usage is logged and the logs are made public.
• This it is hoped that it could be used for
  digital library research.
• Ways will be found to aggregate usage on
  different physical installations.

31
to do list

• finish a version of konz that recognizes HTML
  full text
• integrate DoCIS and DoIS
• finish conversion of DBLP to AMF
• open institutional registration for rclis
• open author registration for rclis
• open a NEP-like service for rclis

32
http//openlib.org/home/krichel
collaboration is welcome!

• Thank you for your attention!