Text Analysis Course

Digital Humanities Master. University of Bern.

Instructor: Elena Spadini.

Content

Date	Topic	Notes	Corpus
18.09	Intro and regex		TGG
25.09	Voyant Tools	With Ursula Loosli (Unibe)	TGG+
2.10	Lexical diversity		TGG+
9.10	Word embeddings		TGG+
16.10	Tokenization, lemmatization, and POS tagging		TGG
23.10	Case study 1: the workflow		Frankestein
30.10	Case study 1: categorising variants		Frankestein
6.11	Network analysis		TGG
13.11	Network analysis	With José Luis Losada Palenzuela (Unibas)	TGG+
20.11	Infoclio Conference (see below)
27.11	Data visualisations	To be rescheduled (see below)
4.12	Case study 2: the workflow		TGG
11.12	Case study 2: gender agency in dialogues		TGG
18.12	Presentations

Calendar

Infoclio Conference 2026 On the 20th of November, we go to the https://www.infoclio.ch/en/programm-tagung2026. Please confirm to me you will be able to attend 10:00-12:00. If you want to stay for the whole day, you will have to pay the registration fee of 15 CHF. This is not mandatory.

Reschedule On the 27th of November, our lesson is cancelled. This lesson is rescheduled to … 23rd or 30th of November ?

Corpus

FITZ-TGG-25 or TGG = Francis Scott Fitzgerald, The Great Gatsby*, 1925 (~47'000 words)

CATH-ALL-23 = Willa Cather, A Lost Lady, 1923 (~37'500 words)

HEMI-TSAR-26 = Ernest Hemingway, The Sun Also Rises, 1926 (~67'500 words)

TGG+ = FITZ-TGG-25 + CATH-ALL-23 + HEMI-TSAR-26.

These texts are available via the Project Gutenberg, e.g. https://www.gutenberg.org/ebooks/64317. We will use the Plain Text format (see Download options).

Frankestein = materials from https://shelleygodwinarchive.org/.

Create a folder data in your computer, where you collect the texts to work with. A similar folder data is available here.

Assignment

For the assignment, you create your own analysis of one or more texts.

Two deliverables:

• An oral presentation, during the last session
• A Jupyter Notebook

Methods: you can choose between

• word embeddings
• network analysis
• a mixed workflow including LLMs and traditional NLP approaches (lexical diversity, lemmatisation, POS tagging).

Corpus: one or more texts of your choice or ask for it.

Let's discuss it before. Write an email or ask for an appointment to briefly present your plan by November 20 at the latest. You should indicate the chosen corpus and method(s).

Using LLMs and AI assistants

You may use LLMs and AI assistants for this course and for the assignment under two conditions:

• Document it. State which tools you used and for which steps.
• Check it. You are responsible for every claim and every reference.

Tools and more

Markdown

• If you have never encountered Markdown before, please follow the tutorial at https://www.markdowntutorial.com/ (use the English version only).

XML

Text and code editor

• Recommended: VSCodium.

Up-to-date browser

• Recommended: Firefox or Chrome/Chromium.

Jupyter Notebooks

• Read the page and then install.

Regular expressions

• Tutorials online, for example https://www.regexone.com/ and https://regexlearn.com/.
• A regex cheat sheet is available at https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Regular_expressions/Cheatsheet.