Attendees
Alan Stanley
Amy Blau
Dana Bronson
David Wilcox
Kun Lin
Noah Smith
Paige Morfitt
Agenda
Topic
Goals and deliverables
What are our goals with regard to metadata remediation and mapping?
What deliverables do we need to produce in order to accomplish these goals?
Requirements, blockers, and outstanding questions
What do we need in order to accomplish our goals?
Are there any blockers that need to be resolved in order to proceed?
Are there outstanding questions that still need to be resolved?
Timeline
What are the decision points and when do they need to happen?
When are the deliverables due?
Wrap-up and next steps
Resources
Project Plan (Draft)
ARMINDA Metadata Mappings
(draft)
Notes
Goals and deliverables
Whitman perspective
Metadata remediation
Bring use of fields and field names into conformance with new collections standard before moving into the new system
In process now and almost complete
Complete in the sense of having identified which fields to remediate - changes have not yet been implemented
Working in spreadsheets
Use IMI remediation to make changes in Islandora 7
Planning on re-ingesting spreadsheets before migration
Got rid of over 100 fields that were no longer needed
Renamed fields to be clearer
Only added a couple new fields
Making sure all dates, names, etc. are standardized
Looking at about a month to complete
Metadata mapping
All fields have a logical equivalent in RDF
Any fields that still need to be removed could be removed during the mapping
Mapping is about 90% done
Newspapers
Need to understand which fields will be required in a paged content model that might not already exist
Sort by date
Working on TWIG templates
Translates flat CSV
Deliverables needed
Alan: Would like to see spreadsheets detailing metadata to go into new system
Requirements, blockers, outstanding questions
Blockers
Repeating metadata fields are a problem in RDF
Some of this is being helped by remediation work
Staging server need to be updated
This is a potential blocker if things go poorly
Outstanding questions
Content models
Some photographs are not in the content model Whitman would like
Basic vs. large image
Content models are tied to file formats and viewers
We can bring over the original files during the migration, create derivatives as appropriate, and associate the right viewer
JPG2000
Not a good native format for Windows machines
JP2s are not a good archival format, kind of a black box
TIFFs and JPGs are better for downloads as well
Whitman has original TIFFs, which can be supplied along with a spreadsheet connecting them to PIDs
Entities
Adding URI-based entities in metadata would be useful to add sooner rather than later
Need to work out fields where entities would be useful
Need to schedule another meeting on entities
Timelines
Remediation: November
Spinning up new production server: January
Add metadata spreadsheets to Google Drive: October
Metadata and content mapping completion: December
Content freeze: January
Action Items
Alan: S
end static IP info to Kun for SSH access
Amy: Add metadata spreadsheets to shared Google Drive folder