TECH NOTES September 23, 2010

printer friendly

September 23, 2010: Tech Meeting

1. Downloading content from & uploading content to Internet Archive

Presentation

1. Determine a best practice for IA identifiers as used by BHL when uploading our own content
   • Need a naming structure
   • Need to respect/honor the local identifier

ACTION ITEM: Mike, Adrian, Noha to describe a naming structure at BoF, send around for discussion & decision

1. Should write out enhancements made to BHL data in a separate file to be stored at IA, like with names.xml
   1. avoids worry about IA overwriting our changes in their standard files
2. BHL to open discussions again with Open Library
   1. Try to determine longevity of project
   2. Where are data stored?
   3. Is their data merged into IA?
      1. We should demo our names.xml strategy for them
   4. Could we take their data and merge into BHL to enhance our own data?

ACTION ITEM: Chris, Martin & Suzanne to talk with George Oates at IA meeting in October

2. GUIDs & Identifiers

• Everyone likes Handles
  • BA using them for every digitized object (each book in a multi-volume series)
• We don't care about assigning ISSNs or ARKs to new digital content
• We're already working on OCLC/ISBNs thru OCLC & shouldn't put special emphasis or work into accomplishing ISBNs (If it works, great, if not, no big loss)
• DOIs are great, but could be expensive

Could we?

1. Assign Handles to all BHL Content
2. Get a single prefix from CrossRef for a Global BHL
3. CrossRef resolves to a Global BHL Handles server
4. Handles server resolves to scanned content

ACTION ITEM: Chris to continue discussions with CrossRef & broach the idea above with them
ACTION ITEM: BHL-EU to build this global resolver / minting system; part of their existing deliverables

3. Name Finding & other services

• Lakshmi worked on NetiNeti
  • http://namefinding.ubio.org/
  • It is AWESOME!
  • provides improvement over existing TaxonFinder
• NetiNeti needs to be backwards compatible with TaxonFinder to keep BHL's existing application going
  • could write a wrapper around NetiNeti to make this happen
  • month or so of work to get this ready for use
• Going forward, it could run on cluster
• NetiNeti could also be used for place names, person names, other entities
  • Needs training data
  • Good project for students / crowd
  • Plug into GoldenGate?

ACTION ITEM: Lakshmi, Ant, Chris, Donat, Guido, Stijn Coleman to discuss possibility of using this with GoldenGate & using GoldenGate output as training set, BHL as testbed
ACTION ITEM: Encourage another Nomina workshop on this topic with GBIF & other names-based services