Open Archives Initiative

1
Open Archives Initiative

• Protocol for Metadata Harvesting

2
Collections in isolation

• Some thoughts
• A wonderful collection is of limited use if it is
  not well known.
• Very redundant collections are often wasteful

3
Virtual collections

• Some collections do not contain actual materials,
  only information about materials and links to the
  home site.
• How do these virtual collections get the
  information about other collections? How do they
  stay up to date?
• --gt The Open Archives Initiative Protocol for
  Metadata Harvesting (OAI-PMH)

4
OAI - PMH

• A protocol -- that is just an agreement to
  exchange messages and interpret them according to
  strict rules.
• Metadata -- data about the data -- information
  about the material in the collection
• Harvesting -- gathering in the desired part of
  the collection for further use

5
The protocol

• See http//www.openarchives.org/OAI/openarchivespr
  otocol.html
• Two sides - the repository and the harvestor
• The repository (data providers)
• Prepares the required metadata
• Responds to the harvester queries
• Acts like a server - responding to queries when
  they come
• The Harvester (data gatherer)
• Gathers the metadata from the collections
• Organizes the harvested metadata in a way to
  serve its purpose.
• Acts like a client - requesting service when it
  needs it.

6
Resource, item, record

• Resource the actual content of the collection
  the point of the digital library
• Item a part of the repository that generates the
  metadata.
• Record metadata in a specific format available
  for dissemination.
• Encoded in XML
• Unique identifier
• Datestamp
• setSpecµ
• Optional status

7
Sets

• Repositories may organize items into sets
• Allows selective harvesting
• Each node in a set organization has
• setSpec
• Set may be hierarchical. If so, the levels are
  separated by colons
• setName
• setDescription

8
Requests

• Request embedded in an HTTP request
• Valid OAI PMH Requests
• GetRecord
• Identify
• ListIdentifiers
• ListMetadataFormats
• ListRecords
• ListSets

9
GetRecord

• Required arguments
• Identifier unique identifier of an item whose
  record is requested
• metadataPrefix prefix part of the metadata
  record relevant to the requested item
• This identifies the type of metadata applied to
  the record. Example oai_dc (the OAI version of
  the Dublin Core -- standard 15 elements, no
  extension.)
• Errors badArgument, cannotDisseminateFormat,
  idDoesNotExist

10
Identify

• No arguments
• Requests information about the repository.
• Response includes
• repositoryName
• BaseURL
• protocolVersion
• earliestDatestamp
• deletedRecord (how does the repository handle
  deletions -- no, transient, persistent
• Granularity (how finely can the datestamp be
  specified?)
• adminEmail
• compression (what schemes are supported)
• description

Optional
11
ListIdentifiers

• Required Argument
• metadataPrefix
• Optional Arguments
• from
• until
• set
• Exclusive argument
• resumptionToken (flow control token for resuming
  an incompleted previous ListIdentifiers request)
• Errors badArgument, badResumptionToken,
  cannotDisseminateFormat, noRecordsMatch,
  noSetHierarchy

12
ListMetadataFormats

• Optional argument
• identifier (if metadataformat is needed only for
  some particular item)
• Errors - badArgument, idDoesNotExist,
  noMetadataFormats
• Response includes both metadataPrefix and the
  associated schema

13
ListRecords

• Required arguments
• metadataPrefix - Only records for which the
  specified metadataPrefix applies should be
  returned
• Optional arguments
• from
• until
• set
• Exclusive arguments
• resumtpionToken

14
ListSets

• Exclusive Argument
• resumptionToken (used to continue a previous
  incomplete response to ListSets)
• Errors - badArgument, badResumtpionToken,
  noSetHierarchy

15
Resources

• Compliance testing -
• www.dlib.vt.edu/projects/OAI/repexp/repexp.html
• OAI PMH -
• www.openarchives.org/OAI/openarchivesprotocol.html
  
• Implementation Guidelines
• www.openarchives.org/OAI/2.0/guidelines.htm