NEH Institute

planning week 2 [suggestion]

Note: each day starts with a recap on modelling, focused on the topic of the day (respectively tokenization, collation, annotation, queries), and ends with time for review.

updated planning week 2

Week 2, Day 2: Tuesday, July 18

Synopsis

Week 2, Day 2 introduces the idea of processing pipelines. It focuses on modular approaches to and algorithmic aspects of textual criticism. Outcome goals:

• A modular understanding of textual criticism
• Modular approaches to designing and implementing an edition

9:00–10:30: Computational pipelines and text models

Introduction to the idea of computational pipelines.

The Gothenburg model (GM) of textual variation can serve as an example of a computational pipeline for the analysis of textual variation. Focus here is not on textual variation: tokenization and normalization are necessary steps for every form of text processing and analysis.

Concepts:

• Tokenization
• Normalization

11:00–12:30: Tokenization 1

Tokenizing using different text models: plain text, TEI/XML.

"Simple" tokenization: plain text

• Exercises
• Materials from the Amsterdam workshop?

2:00–3:30: Normalization

Normalization strategies

• part of speech tagging
• upper case / lower case normalization
• etc. Issues Materials(?)

4:00-5:30: Review

Review section. Focus on participant's questions.

Alternative (in case all is clear) is to focus on markup as an expression of a data model. Expand upon the tokenization different kinds of text with markup (TEI) or, more generally, processing text models.

Week 2, Day 3: Wednesday, July 19

9:00–10:30: Modelling (collation)

Modelling: collation, transcription, and markup. Concepts.

11:00–12:30: Tokenization 2

Advanced tokenization: XML markup; Unicode; SoundEx.

2:00–3:30: Collation 1

Collation.

• Gothenburg Model (GM)
• Theory of automated collation and textual variation analysis
• Collation goals
• Exercises with automated collation software: CollateX and Stemmaweb?
• near-matching?

4:00-5:30: Collation 2

Collation and TEI/XML markup.

• different approaches to markup collation:
  • flattening text
  • passing along markup
  • JSON (Sydney workshop)

Week 2, Day 4: Thursday, July 20

Modelling: Annotations, Transcription. The goal is to make participants aware of "Research-driven annotation", letting them ask questions: 1) What are the inherent properties of the text 2) What do I need for my research? Discribe the computational pipeline: research questions → data model (including query facilities) → markup/annotation.

9:00–10:30

• Definition of Text Annotation
• Pipeline of making annotations to text
• Different approaches to and forms of annotation (allude e.g. to Alexandria)
  • TEI
  • Text-to-image linking; IIIF
  • etc.

11:00–12:30

...

2:00–3:30: Text analytics 1

MK: Bag of words, text processing, text as tables, query the tables

4:00-5:30: Text analytics 2

MK: Bag of words, text processing, text as tables, query the tables (continued)

Week 2, Day 5: Friday, July 21

9:00–10:30

11:00–12:30

2:00–3:30: Text analytics 3

MK: Unsupervised learning, cluster analysis, PCA, paleographic analysis

4:00-5:30: Text analytics 4

MK: Unsupervised learning, cluster analysis, PCA, paleographic analysis (continued)