Establishing National Digital Repository System employing Harvesting Model

1
Establishing National Digital Repository System
employing Harvesting Model

• Surinder Kumar
• Technical Director, NIC, New Delhi
• suri_at_nic.in, 011-24305503

2
IRscontd

• At present, the University of Southamptons
  worldwide registry of OAI compliant open access
  repositories lists more than 1000 repositories.
  Number of IRs produced by India is around 50. To
  make it available as single virtual archive and
  also means of providing seamless search, it is
  becoming essential to form a network of connected
  research repositories and resource discovery
  services to form National digital repository
  system. Examples are CARL, ARROW, DRIVER etc

3
National Digital Repository System

• To build an appropriate NDRS, analysis of
  existing infrastructure are analyzed.
• Technology Components
• Requisite Hardware
• OS
• IRs software such as DSpace, Eprints
• Interoperability among IRs is proven with the
  development of OAI-PMH protocol by OAI.

4
Technical Model of NDRS

• Alma Swan and Chris Awre has mentioned three
  models in Linking UK Repositories. These are
• Centralized Model
• Distributed Model
• Harvesting Model

5
Centralized Model

• metadata and content are submitted directly to a
  central server.
• Advantages
• Have complete control of the whole process from
  article deposition through to the user interface
• Software selection
• Able to manage preservation issue
• Disadvantages
• It is an expensive option
• It may surpass the existing institutional
  repositories

6
Distributed Model

• All metadata and content remain in their source
  locations and metadata is searched on the fly.
• Advantages
• providing up-to-date metadata as it provides
  instant access to source locations of metadata
• Relatively very less expensive as compared to
  centralized model
• Disadvantages
• No enhancement of metadata
• Network dependent
• Not many IRs support Z39.50 or SRU/W

7
Harvesting Model

• It is a hybrid model where metadata is harvested
  into a central searchable server and also
  distributed as content (full text) would be
  provided by individual repositories. Under this
  model, service provider would harvest metadata
  from existing institutional repositories using
  the Open Archives Initiatives Protocol for
  Metadata harvesting (OAI-PMH). Service provider
  can enhanced the quality of metadata and provide
  the various services from their centralized
  server. The metadata canbe further exposed via
  OAI_PMH, SRU/W, RSS feed for use by other service
  providers.

8
Harvesting Model-advantages

• Advantages
• OAI-PMH is a standard protocol which is easy to
  implement
• Unqualified Dublin Core is mandated to be
  OAI-compliant, however, more complex metadata
  schemas can be employed.
• The institutional archives employ software which
  supports OAI-PMH
• Harvesting can be carried out by automatic
  scheduled tasks

9
Harvesting Model-disadvantages

• Only Unqualified dublin core is mandated for
  harvesting, it lacks rich semantic as compared to
  other metadata schema
• The metadata exposed by the services may not
  always latest. Also changes made in metadata may
  not be reflected in the central server.

10
NDRS-Accepted Harvesting Model

• It is clear that OAI-PMH model has much
  advantages as compared to other model
• It has gained worldwide acceptance
• It makes easy to share information about
  scholarly resources and to offer enhanced
  resource discovery tools.
• It has been adopted by thousands of institutions
  around the world.

11
NDRS-benefits

• National Digital Repository system would offer
  number of benefits to end users as well to the
  various stake holders of the Institutions.
• Benefits to IR Administrator
• IR administrator would only maintain the content
  of the repository while offering metadata to
  service provider.
• NDRS would be inbetter position to provide long
  term preservation through appropriate metadata
  provision and/or content package
• It would offer an enhanced metadata to the end
  users

12
NDRS-benefitscontd

• End Users as readers and searchers
• NDRS would provide end users access to a large
  number of repositories rather than accessing
  individual repository.
• It would push the content to end users through
  RSS/ATOM feed.
• It would provide document delivery services to
  the end users

13
NDRS-benefitscontd

• End Users as a content manager
• NDRS would provide means to expose authors work
  so as to make their work widely available to
  their peers throughout the globe.
• It would able to provode provide preservation and
  metadata enhancement capabilities to support the
  long term storage and access to the content.

14
NDRS-benefitscontd

• Content Aggregators
• NDRS would offer added-value services of their
  own to enhance aggregated metadata and supply
  this back to the repository concerned.
• IT would provide a single point of information
  for statistics about access and downloads of
  data.
• It would offer a single point of information to
  multiple source of research and other materials
  to aid discovery.
• It would able to provide certain collections by
  adding value added services on top of it.

15
Impediments in implementing in NDRS

• Technical issues at data provider levels such as
  installation of IR software, server, server
  malfunctioning, backup of data and updating of IR
  software etc whereas in case of service provider
  level, successful harvesting of data involves
  error free network, the proper use of Dublin
  core metadata field, data sets and problems with
  the correct use of date stamp etc.
• Coordination among IR members
• Federated Authentication and Authorization
• Long term preservation, format, migration and
  access
• Sustainability in providing ling term access to
  NDRS

16
Current Scenarios of Institutional Repositories
in India

• Registry of Open Access Repositories (ROAR) lists
  52 repositories have been registered, however,
  this number may be higher as certain repositories
  have yet not been registered with ROAR.
• Analysis of IRs in India
• Out of 52, 13 were not functional at the time of
  writing paper
• Number of them have not been updating
• To look further, it is not reaching the critical
  mass

17
Current Scenario..contd

• As per survey conduced by Webometrics 2010 for
  latest ranking of Worlds open access
  repositories for visibilities, quality and
  available items18, there are seven repositories
  listed from India and their details as given in
  the following table.

18
Sr No. Rank Name of IR Year of establishment No of records
1 82 Indian Institute of Science 05-04-2004 19477
2 148 OpenMed, National Informatics Centre 22-03-2005 2645
3 180 Indian Statistical Institute digital Library 17-01-2004 188
4 218 Indian Institute of Astrophysics 11-11-2004 2468
5 245 National Institute of Oceanography Digital library 06-04-2010 3528
6 278 Raman Research Institute Digital Library 19-04-2005 3731
6 278 National Aerospace Laboratories Institutional Repository 9-11-2004 3164
19
Current Scenario-service providers

• There are 9 service providers in the country who
  are harvesting data majority of them follows
  OAI-PMH and harvesting software used is PKP
  Harvester. Out of 9, four are not functional,
  though these are highly cited in the literature.

20
Proposed NDRS

• Establishing successful, well populated National
  level repositories, we need to look at prevailing
  information system in our country. For example,
  ICMR, CSIR, ICAR, Envis, Deptt of Atomic Energy,
  ISRO. Onus should be on those national
  information system should able to provide
  publications arising out of public funded
  research should make it available free of cost to
  researchers

21
(No Transcript)
22
NDRS-Recommendations

• There is a need of national body in the country
  as in JISC in UK who is providing advisory as
  well technical services to individual
  repositories
• Responsibility should be given to National level
  organizations to set up a national resource
  centre that should harvest data from their
  respective institutional repositories
• Develop strategies to make institutional
  repositories a permanent and sustainable part of
  the national and local research infrastructure
• Guidelines to the respective institutional
  members mediate deposit or voluntary deposit and
  needs for mandatory deposit of papers and
  dissertation
• Develop guidelines for metadata entry and best
  practices followed

23
Conclusion

• There is a new challenge to create an environment
  based on OAI protocol so that public funded
  research should be made available to the whole
  community
• National level body is needed so that development
  in institutional repositories should be more
  coherent as it may able to provide the best
  advisory services and adoption of guidelines set
  and best practices followed by various national
  level systems such as DRIVER, DAREnet, HAL

24

Thanks