Leaving #Seoul after an intense week!

The #ICLA2025 Congress was a blast. I especially enjoyed our "Digital Comparative Literature" sessions; great discussions! And the round table on responsible use of generative LLMs turned out really great as well, with a rather large audience.

The only less-than-great thing was that, unsurprisingly, I think I was basically the only one posting about the event here, with @icla_dcl / @rebsim.

#DigitalHumanities #Comparative Literature #CLS

#NowPlaying on #BBC6Music's #6MusicsRaveForever

CLS & Todd Terry:
Can You Feel It (Intro)

#CLS #ToddTerry

https://toolroom.bandcamp.com/track/can-you-feel-it-montel-remix-2

Letzte Woche auf der #DH2025 habe ich unseren interdisziplinären Ansatz präsentiert, der qualitativ-philologische Verfahren mit NLP-basierten Analysen kombiniert, um sprachlich-linguistische Muster als Heuristiken für kulturelle, politische und ästhetische Semantiken in serieller Literatur im Kontext historischer „Frauen-Zeitungen“ zu nutzen.

Vielen Dank für die anregenden Fragen und Gedanken!

Today’s #highlight for me on the last day of #DH2025? Luca Giovannini presenting his "Modest Proposal for Operationalizing Dramatic Texts". He proposes drama vectorization not to simplify plays, but rather to continue formalist morphological thinking using computational tools. #CLS #drama

What are the key components of a #play?
#DH2025: Luca Giovannini takes us on a tour through centuries of thinking from #Aristotle’s #Poetics, through Russian formalism, to modern literary theory and unpacks the question offering a fresh perspective.
#drama #operationalization #CLS

#DH2025: What are we really measuring when we count words? Artjoms Šeļa, Thomas Haider & Petr Plecháč show how different meters distinctly shape seemingly unrelated feature distributions across Czech, German & Russian #poetry. Their main finding: Mind the meter! #CLS

#DH2025: Steffen Pielström, Kerstin Jung & Patrick Helling put the SPP #CLS projects to the test, examining their level of #FAIRness as a case study to assess how equal automated FAIRness tests are.
#FAIR #FAIRprinciples

It’s already the last day of #DH2025!
Among today’s posters, don’t miss the contribution by Evgeniia Fileva, Christof Schöch, Julia Dudar and colleagues:
“CLS INFRA: Leveraging Literary Methods for FAIR(er) Science.”

We’re looking forward to another day of inspiring conversations

#DH2025: With their research question, @axelpichler &
@janispagel hit the nail on the head :
Can #LLMs apply redefined literary concepts in the prompt-and-predict paradigm (Liu et al. 2023)?
#CLS #concepts #plasticity

#DH2025: Xiaoyan Yang & @fpianz present a #multilingual #CLS pipeline comparing #English & #Chinese #fanfiction. Using #BookNLP, #ARMCC & #LLMs, they address #CoreferenceResolution, dialogue speaker & character feature extraction hot topics at this year's conference.

#DH2025: Another #alignment project in #CLS. Anastasia Glawion and Dhara Lechner of DHSS FAU Erlangen-Nürnberg present their work on "Tracing #Transformation: editorial shifts in the Grimm brothers’ tales".
#FairyTales #German-language #fiction

Neat idea by Jessica Monaco and Mark Algee-Hewitt, at #DH2025, on subgenre mixture within the Gothic novel:

First, train a model to classify text segments into one of eight genres, not including the Gothic (!).

Then, ask the model to classify segments from Gothic novels into those genres in order to obtain characteristic genre mixtures within the Gothic novel.

They obtain radar plots for each Gothic novel that show this mixtures.

Excited for the next session: "Digital Methods for the Study of Theatre, Drama and Performance" Really looking forward to the talks and discussions!

#DH2025: #CLS is everywhere! In SP-09, Allison Anne Keith, Antonio Rojas Castro, et al. present their approach to using #ComputationalMethods to examine, how annotated #gracioso characters can help us understand their structural and narrative role in #Calderón's plays.
#Spanish #drama #annotation

#DH2025: So exciting to see #hallucination rates calculated in #CLS with #LLMs. Great work by Sarah Griebel, Layne-Worthey, et al. on the ongoing challenge of #CharacterAttribution in English fiction titled "Strictly Speaking: Character Attribution in Literary #Dialogue with Language Models"
#SpeakerAttribution

#DH2025: So exciting to see #hallucination rates calculated in #CLS with #LLMs. Great work by @sgriebel.bsky.social, Layne-Worthey, et al. on the ongoing challenge of #CharacterAttribution in English fiction titled "Strictly Speaking: Character Attribution in Literary Dialogue with Language Models"