Agenda:

Conversion Code Update / Notes Testing

Curation Concerns code up and running at: http://mods2rdf.xyz/

Works now in an alpha form for title, abstract, genre, and subject topics.

Steps for usage:

Create an account on the site
Go to http://mods2rdf.xyz/uploads and upload a MODS XML document.
To view the end result, you can do a normal search (but currently on uses fields CurationConcerns knows about). Otherwise, to view in fedora, the credentials are:

username: darthvader
password: tapdancing

Still working on testing performance of storing strings vs. minting objects
Email Steven with any questions

Collaborative Subject Mapping Document (https://docs.google.com/spreadsheets/d/1_YwNyRkegT9gCoIwbZAdOupEeMsCpC9FQ0BXX4E1p-A/edit?usp=sharing)

Mints an object whether there's a URI or not, partly to support SPARQL endpoints, makes labels available/searchable through SPARQL. Other advantage is ability to expand upon term over time, add more links to other terms, etc.
Eben put question out to Code4Lib list to ask about using combination of literals and objects and the one person to respond recommended against it
Will assume that similar strings are the same
Allows for updating labels for every instance in records
Adds other proprties pointing to subelement mappings, ex. dcterms:spatial for geographic subheading pointing to respective subheading in another authority
Question about spatial object properties: if you have an exactMatch to an object with coordinates, is it overkill to include your own coordinates? (ex. DCE:coverage "east=42...)
For local complex terms or complex headings where only subheadings have corresponding URIs, proposed to include rdfs:label for each subheading (but not for complex headings with URI matches?)
Question: what will the workflow be for creating new subject entries? A: could be autocomplete with existing terms or create a new subject object by entering term (and trying to match?) or URI and pulling labels from matched URI.
Things not mapped: temporal (edm:timeSpan)
NYPL looking at option to use combo of edm:timeSpan, PeriodO (http://perio.do/ ), other vocabs with dcterms:temporal property
Hierarchical geographic: a bad idea to try to represent entire hierarchy? You really just need most specific point. How to represent specific addresses? Spatial object would have to have some way to represent full address in addition to coordinates
For next meeting, flesh out examples for these geographic and temporal elements

Classification Individual Institution Mappings (https://wiki.duraspace.org/display/hydra/Classification+Individual+Institution+Usage+And+RDF+Conversion)

Only BPL and Columbia supplied mappings, but we will create collaboration document

Next meeting:

June 13 (skipping Memorial Day)
<identifier> mappings
Review subject modeling for temporal, scale, projection

Browser not supported