Creating a Digital Edition with Generative AI
in a Seminar Setting

Ulrike Henny-Krahmer (Universität Rostock)

TEI Conference, Vancouver, August 13, 2026.

Presentation: https://hennyu.github.io/tei_26/

Overview

Digital Scholarly Editions, Generative AI, and DH teaching and learning a group of people sitting together in a circle for learning
The seminar: Sophie Tiecks Dramatische Fantasien. Workflows für eine digitale Edition mit generativer KI a portrait of Sophie Tieck

Digital Scholarly Editions, Generative AI, and DH teaching and learning

DSE and generative AI

  • typical steps in a digital scholarly edition workflow:
    • text transcription
    • text encoding
    • text conversion
  • GenAI can support these and other steps in a DSE workflow (see, e.g., experiments documented in Pollin et al. 2025)

Some reasons for using GenAI in teaching and learning about DSE

  • current ubiquity of GenAI (relevance for job market and to students' daily lives)
  • possiblity to combine GenAI topics with other DH topics (hence, also DSE)
  • methods easily accessible to users from a single platform via conversation with the bot
  • reflective aspect is essential

The seminar: Sophie Tiecks Dramatische Fantasien. Workflows für eine digitale Edition mit generativer KI

Practical information about the seminar

icon clock
Clock icons created by Magnific - Flaticon 		Duration: April to July 2026, weekly, 13 sessions of 1,5 hours each
networking icon
Team icons created by Magnific - Flaticon 		Participants: 9 students of German Studies (Bachelor or Teacher education)
group room icon
Meeting room icons created by cube29 - Flaticon 		Room: traditional seminar room with screen, whiteboard, flip chart, eduroam

Goals of the seminar

  1. learn about the female German writer Sophie Tieck (*1775, †1833)
  2. create a DSE prototype of one of her Romantic plays (Frühlingszauber, 1804)
  3. thereby learn:
    • what a DSE is
    • about DSE workflows (steps, collaborative work)
    • about GenAI
  4. reflect on the usage of GenAI for scholarly tasks

The play Frühlingszauber

a page of the play Frühlingszauber another page of the play Frühlingszauber

Workflow and tools used in the seminar

  • GitLab repository (hosted by University of Rostock)
  • Visual Studio Code Editor (hosted by University of Greifswald)
  • Chatbot (gemma4, hosted by University of Greifswald)
  • Browser (to look at HTML results)
 workflow diagram

Step 1: text recognition

  • zero shot prompt with role, instructions, goal, constraints
  • text was extracted from PDF of images that contained basic OCR
  • problem: chatbot did not process the whole file at once

Step 2: text encoding

  • a very long prompt (one shot) explaining the TEI encoding strategy and giving one example
  • this worked very well!
 TEI encoding example

Step 3: text conversion

screenshot of the homepage
  • individual prompts to produce HTML (with CSS and JS), directly from TEI
  • astonishing results were produced

Evaluation of the seminar by the students (1)

rating of AI usage

Evaluation of the seminar by the students (2)

advantages of AI usage

Evaluation of the seminar by the students (3)

drawbacks of AI usage

Evaluation of the seminar by the students (4)

ranking of chatbot usage

Thank you very much for your attention!

References

  • Pollin, Christopher, Franz Fischer, Patrick Sahle, Martina Scholger, and Georg Vogeler (2025): “When it was 2024 – Generative AI in the Field of Digital Scholarly Editions.” Zeitschrift für digitale Geisteswissenschaften 10. https://doi.org/10.17175/2025_008.
  • Fritsche, Katrin und Sander Münster (2024): “Taking up Artificial Intelligence as Teaching and Learning Content in the Digital Humanities – Topics, Categorisations, and Examples.” In Proceedings of the 8th International Conference on Advanced Research in Teaching and Education. https://doi.org/10.33422/icate.v1i1.225.
  • Haberstok, Monika (ed.) (n.d.). Sophie Tieck [Website]. https://www.sophie-tieck.de/.

Presentation:
https://hennyu.github.io/tei_26/

CC-BY 4.0