Mapping myExperiment.org Packs to the DataBank Repository

I. Problem Summary 

The impetus for the internship was the user-driven need for a mechanism with which to archive and reference Packs in scholarly communication.

Pack to citable object

http://www.doi.org/logo.html

II. Initial Proposals

Research Object Principles

http://www.researchobject.org/overview/

Pack Formalization

• Packs to be formalized into Research Objects
• Support for this is being developed in myExperiment (Alpha) implementation of the RO API
• May also be formed manually using the RO Manager console application

Deposit in DataBank

• Existing Support for RO Ingest
  • RDF Manifests should be mergable
  • Compressed directories retain structure when unpacked in DataBank
• Bottlenecks
  • Databank is unaware of structure of objects
  • Only unpacks top level of files

Unique Identification

• Value of issuing a DOI
• Existing DOIs wrapped in the RO should be documented in the manifest
• DOI will resolve to the DataBank landing page for the RO

III.  Summary of Work Completed

Interviewed Stakeholders

• Oxford e-Research Centre myExperiment team
• Bodleian Digital Library Systems and Services 
• Research Object model developer

Assessed RO Deposit Workflows

• Current DataBank web interface
• Merging manifests
• Exposing structural metadata

Developed Recommendations

• Research Object final report

IV. Recommendations

Finzalize myExperiment Alpha2 Implementation of RO API 

• Successfully produce ROs from Packs

Improve Discoverability

• Develop customized index from RO manifests
• Extract metadata from nested manifests for indexing

http://www.realmdigital.co.za/post/bridging-the-online-book-discoverability-gap-part-/

Pre-Processing ROs

• Processing on myExperiment end for DataBank submission information package
• Unpack zip files
• Update manifest with new metadata
  • Including existing DOIs or other unique identifiers

V. Proposed Lifecycle

Pack is in Archived state

Eelke van der Horst http://alpha2.myexperiment.org/packs/559

User Initiates Deposit

RO Generated

RO Pre-Proccessed 

• RO pre-processed by myExperiment interface to utilize DataBank API calls

http://www.xtremeesolutions.com/xappso-process/

Processed RO Deposited in DataBank

• DOI issued at time of ingestion

Manifest Indexed 

VI. Future Work

Short Term: External Resources

• Archiving associated external objects
• Attribution & copyright challenges
• Will it still be verifyable, reproducible?

Long term: Provenance

• Capture and embed provenance metadata in archived RO
• W3C PROV Model may be a useful point of entry