Meeting Notes

2022-02-01

Cliff

Setting ground work with python classes and will soon be able to process large quantities of pages to see what we get
Created github project: https://github.com/pulibrary/finding_aid_enrichment
- Information space for this project
- Add standup notes here? Yes.
Models under each use case
- Currently Issues - Move into wiki?
- Actively engage with ADAPT to edit and grow to understand what this is for
First ticket in progress is plugging several existing technologies together.
- SpaCy: Where do strings occur and where do named entities occur?
- - Python data structures

Alexis

Adding Will Clements to this Project
- Will will be invited to future meetings
- Both will review the github repository as he will probably work on some of this and is interested in Data Problems
- Alexis will talk to will this week and will loop him in next week. They will review these.

Esmé

Named entity recognition and library data - CDH folks interested.
Esme will schedule to meet with student to get general sense as to what they want.
- Esme: Needs to figure out how to extract corpus from figgy that they could work with. Plain text or HOCR, etc…
- Cliff: would like to be a part of these meetings and potentially direct them to papers of princeton or blue mountain project if appropriate
- Cliff: Would not direct them to the dirty ocr
- Kim: Student internship program project to review results/work with rough ocr?
- Alexis: Might be worth investigating

Questions

Esme: Has questions about the lower set of tickets but the upper two are fine.
- Cliff: the lower set may well be out of scope/in need of editing.
Alexis: Is there anything that she can read that she can better understand named entity recognition and what it can/cannot do.
- Cliff will compile a list of reading
Alexis: is it ok to adjust or change things in the wiki/github issues?
- Cliff: yes, absolutely it is expected and very much appreciated. Dialogue between Alexis’s group and IT as tech to think through essential.

Action Items

Alexis inviting Will to future meetings
Esmé reaching out to studentand setting up time for Student/Esmé/Cliff to meet (issue #4)
Cliff compiling a list of reading material for Alexis (issue #5)
Kim adding notes to github project wiki (done)

Present

Cliff Wulfman, Esmé Cowles, Alexis, Antracoli, Will Clements, Kim Leaman (notes)

Esmé and Cliff

Met with Student, and recommended the Daily Princetonian Archive to be used for the student’s project.

Alexis

added a hypothesis that is not entirely clear, but worth discussing
- While doing this work and who is named/what name is in each component, can that lead to manual work between people or places,etc…
- - Esme: yes, and of particular interest would be people who crosswalk between collections

Cliff

Started page using named entity recognition and links to what it is and how we may be using it for this particular project
Working on constructing a robust architecture for taking a iiif manifest, pulling down images associated with each page, cleaning and running ocr on it and initial named entity pass. Collecting those up at a page at a time and into a collection
- Wants to be able to track where names occur- what page, etc.
- Researching how to integrate into iif….RDF - annotations on a canvas?
- Build up a graph of all the strings and locating where they are on each canvas.
- Examine what those strings are, where they occur, etc…
- Take a container/folder or two from the selected materials and see what is coming out of it.

Esmé

Cliff, can you share with everyone on slack what your results are before the next meeting? It would be a good step for folks to review and discuss next week
- Yes

Action Items

Cliff: will share results with folks in Slack to review before meeting next Tuesday
Alexis and Will: Will take a look at the Named Entity link(s) Cliff has shared.
Alexis, Will, Esmé, Kim: Will take a look at what cliff shares in slack.
Kim: add notes to Github