OAI-PMH at Yale Report on the DLF OAI

OAI-PMH at Yale Report on the DLF OAI Training Session November 10, 2005 Charlottesville, VA

Overview • • Review of the protocol OAI best practices Potential Yale applications Next steps for the Metadata Committee

OAI-PMH v. 2. 0 Basic Concepts • Data provider: administers systems that expose metadata • Service provider: uses metadata to build value-added services • Harvester: a client application that issues OAI-PMH requests • Repository: a network accessible server that can process OAI-PMH requests

OAI-PMH v. 2. 0 Basic Concepts • Resource: the physical or digital object that metadata is "about" • Item: a constituent of a repository from which metadata about a resource can be disseminated • Record: metadata in a specific format • Identifier: a unique identifier that unambiguously identifies an item in a repository; must conform to URI syntax

OAI-PMH v. 2. 0 Harvesting • Deleted records • Sets • Datestamps – ISO 8601 – UTC • Selective harvesting

OAI-PMH v. 2. 0 Protocol Features: HTTP • Request – GET base. URL? key=value&…. &key=value – POST base. URL Content-Type: application/x-www-form-urlencoded Content-Length: number of characters key=value&…&key=value • Response – XML document in message body or error code

OAI-PMH v. 2. 0 Protocol Features: XML Response • XML declaration • OAI-PMH root element with these attributes: – Default namespace declaration xmlns=“http: //www. openarchives. org/OAI/2. 0/” – Schema instance declaration xmlns: xsi=“http: //www. w 3. org/2001/XMLSchema-instance” – Schema location xsi: schema. Location=“http: //www. openarchives. org/OAI/2. 0/OAI-PMH. xsd”

OAI-PMH v. 2. 0 Protocol Features: XML Response • response. Date element – YYYY-MM-DDThh: mm: ss. Z • request element base. URL • response element – It has the same name as the verb used in the request.

OAI-PMH v. 2. 0 Protocol Features • Multiple metadata formats – metadata. Prefix • Flow control – resumption. Token

OAI-PMH v. 2. 0 Requests & Responses • Get. Record • Identify • List. Identifiers • List. Metadata. Formats • List. Records • List. Sets

Current Work: Resource Harvesting within the OAI-PMH Framework • Datestamps – Updated record vs. updated resource • Locating the resource – Multiple URLs: splash page, resource, etc. – Multiple elements used inconsistently: dc. identifier, dc. format, dc. relation

Current Work: Resource Harvesting within the OAI-PMH Framework • Complex object formats – FOXML – METS – MPEG-21 DID – SCORM • Other implementations – mod_oai

OAI Best Practices DLF OAI Implementers Workshop Handouts from the session 1. Project Abstract 2. The Case for OAI 3. OAI “Cheat Sheet”: A Taxonomy of Rapid OAI Deployment Strategies 4. Summary of OAI Metadata Best Practices 5. Summary of the DLF Aquifer MODS Profile 6. OAI Tools 7. OAI Implementation: Administrative Planning

OAI Best Practices Implementation Decisions • Collections – Develop criteria. Prioritize according to ease of implementation, associated risk, logical dependencies among items, etc. • Metadata formats – Decide which formats to support. • Technical infrastructure – E. g. , use a gateway that provides a base URL for multiple individual collections.

OAI Best Practices Deployment Options • Emory’s Metadata Migrator • Static repositories • UIUC’s OAI File. Maker. Pro Gateway • Fedora • Luna Insight

OAI Best Practices for Data Providers • Identifiers – Should be persistent & unique. – Should not be reused. – Specification and XML Schema • Datestamps – Use UTC. – Support seconds granularity, if possible. • Deleted records – Provide persistent support, if possible.

OAI Best Practices for Data Providers • Resumption tokens – For repositories > 2 MB • Sets – Service providers harvest by set. – How should sets be organized? • About containers – Rights – Provenance (for 3 rd party aggregators)

Implementation Guidelines Includes: Guidelines for Repository Implementers Guidelines for Harvester Implementers

OAI Validation • Reap: OAI command line harvesting • Repository explorer: for data providers & service providers to test harvesting & searching • W 3 C validator for XML schema • Utf 8 conditioner: for character encoding problems • See OAI Tools handout for more info.

OAI Best Practices for Shareable Metadata The four C’s of shareable metadata • Consistency • Coherence • Context • Conformance

OAI Best Practices for Shareable Metadata • Metadata in a shared environment – Context & coherence – Don’t assume a local user. • Granularity of description – Appropriate for access to the resource – Don’t expose records for subordinate items. • Use of multiple metadata formats – Need to be expressed as XML schema – Stepped crosswalking to simpler formats.

OAI Best Practices for Shareable Metadata • Relating versions of a resource – One-to-One Principle – Multiple strategies/compromises • Document metadata creation practices – In OAI responses – In external documentation • Communication with service providers

Potential Applications at Yale Implementation Goals • Improve user experience – Federated search • Improve management of resources – Finding aids • Collaborate with institutional partners – AMEEL • Develop digital library infrastructure – At Yale and beyond

Potential Applications at Yale Resources & Roles • Resources – Commitment of stakeholders – Analysis of deployment options – Server infrastructure – Staff hours • Roles – OAI-PMH Implementation Manager – Programmers & technical staff – Metadata specialists – Digital collection curators

Potential Applications at Yale Sharing Metadata • 3 rd Party Aggregators – OAIster – DLF Portal – MODS Portal • Registries – Registered OAI repositories – Institutional Archives Registry – OAI Registry at UIUC

Next Steps for the Metadata Committee • Centralized implementation at Yale? If yes, – Relate to other digital library initiatives. – Create buy-in. • Service provider needs – Consult with IAC committees. • Data provider needs – Consult with digital collection curators.

Next Steps for the Metadata Committee • Metadata recommendations – Recommend multiple formats – Decide upon a common format • YES? MODS? • Stepped crosswalking from other formats – Content & encoding guidelines – Metadata creation tools – Staffing