TechCall_30jan2017

Agenda
1) Review the BHL Egypt list of work packages

https://docs.google.com/document/d/1-12MJCYcWBPcR-NzBri8Lx60whTMHNYQO1TEkhTpJWo/edit?usp=sharing

2) Review incremental changes submitted via Gemini

3) (OPTIONAL... there is already an email thread about this) Determine requirements for how to submit articles for DOIs. The email contents are:

Right now, the settings for the service allow me to:

a) Submit (or not) titles for DOI assignment. Up to 10000 titles without DOIs are submitted at once.
b) Submit (or not) articles for DOI assignment. Up to 10000 articles without DOIs are submitted at once.
c) The number of entities submitted at one time (i.e. 10000) is also configurable.

What do we want to add to further limit how DOIs are assigned? There has never been a need to do anything different with titles, so I am assuming nothing relating to DOIs and titles needs to be changed. For articles, do we want to be able to limit DOI assignment to a journal (every article associated with every item associated with the title/journal), or to a volume of a journal (every article associated with the item/volume)? Or either a title/journal or an item/volume? Or is there some other way entirely that we would like to control exactly which articles are submitted?


NOTES
Reviewed most recent Gemini issues and updated as necessary.

Server status - lost in the mail. Company is going to send another one.

Year metadata in Macaw: year and date will show up verbatim.
Could we enforce formatting? That is not currently happening.
This issue is part of the year cleanup we discussed the other week.
Rules for ingest will try to force it into the format we discussed.
Will review Macaw to get a sense of how messy the data can be for this as it's entered by those using Macaw.

Susan: We’re negotiating with DPLA and passing them metadata - should they ingest at title level or as item record? Settled on item record; Year is really important. Four digit year or - two digit month - two digit day...
Even one item might span 1800-1802

User created article metadata
Gemini ticket regarding re-use of article metadata collected when someone creates a PDF; awhile back (2011?) Trish and Bianca did a report on this and found most of the user-supplied metadata was usable/useful.
We'll want to investigate and plan on how best we can re-use that stuff. It's relatively good data so we could be using it elsewhere. In particular, it could be used to create a segment.

Some alternatives - opened a Gemini ticket a few months ago; we could crowdsource article definition; There could already be a platform that we could re-purpose for something like this. Maybe wedigbio or zooniverse? There would be people who really care about this content and would want to define the metadata.

DOI assignments

Martin would like to establish a workflow where he can pick a journal title and then request that we assign DOIs to all articles within that title, including within preceding titles.
Could start by finding one title ID, and then any other Title IDs associated with it.
Person may have to submit request, perhaps Bianca would be the best person. They would have to tell Mike that here’s the title, all of the preceding titles; give that in a package.

We may need/want some sort of moving wall on assigning DOIs.
i.e., Assign DOIs to all articles in Volumes 1-10, or 50-80

We would want to check CrossRef first; i.e. via JSTOR
Submits metadata to crossref, if crossref doesn’t find any DOIs associated with that metadata, it goes ahead with assigning a DOI, or adds an existing one to our record if it finds a match.
Accuracy seems to be better with articles than with monographs.

Rolling article metadata definition
Vol 1 , define half now
defines additional for another 30% at a later date

is there a way to get a report for articles that don’t have DOIs?
Could do it periodically
and then generate DOIs for those
or could check every time process runs

would also give us a list of all titles we’re assigning DOIs to

Table of titles
Would we still want to limit to certain volumes?
yes

emails lately mostly between NYBG and Harvard Botany
Defining article definition process to define an individual letter in archival correspondence makes sense.
Setting up as an article part segment
Diane Rielinger

BHL Egypt
Keep building the list
We’ll review next week to ensure the items are what we really want