As always: #OpenData persistently available at:
Du, K. (2025). Reconstructing Shuffled Text (Derived Text Formats) [Data set]. Zenodo. doi.org/10.5281/zeno...
#CLS #CCLS25 #DTF #LiteraryComputing #LLM #Memorization

The authors tackle a major challenge in #NLP: #LLM-Memorization and #Copyright
👉 Can derived text formats (DTFs) be used safely for research on #in-copyright texts without enabling reconstruction of the original?
#CLS #NLG #DTF #LiteraryComputing #CCLS25 #OpenScience
jcls.io/issue/118/in...

As always: #OpenData persistently available at: doi.org/10.5281/zenodo.17552902
And the article: doi.org/10.48694/jcls.4215

They offer a critical #survey of the #SoA of #EventDetection in journalism, history, and literary studies. By comparing their model to a storyline analysis framework used in news, they show how fiction and non-fiction can be analyzed studying narrative progression across domains. 📖✨ #CCLS25

As always, #OpenData and #OpenCode are persistently available:
Havrylash, J., & Schöch, C. (2025). Syntetic texts evaluation with #pydistinto. Zenodo. 10.5281/zenodo.15525428.
And the article: doi.org/10.48694/jcl...
#JCLS #CCLS205 #LiteraryComputing #NLG #Evaluation

By inserting an artificial word with precisely defined frequency and #dispersion, they test how well different measures detect what’s truly #distinctive.

Their findings uncover that a #TF-IDF -based measure is more sensitive to dispersion variations than other dispersion-based measures. #Evaluation

As always: #OpenData and #OpenCode persistently available at: doi.org/10.5683/SP3/....
And the article: doi.org/10.48694/jcl...

By randomizing #LLM prompts and analyzing moral #keywords via co-occurrence #networks and hierarchical clustering, @andrewpiper.bsky.social uncovers latent “moral communities” across 20th–21st century #English-language #fiction.

As always: #OpenData and #OpenCode
Keith, A. (2025). calderon-gender-prediction. Archived GitHub Repository. Zenodo. doi.org/10.5281/zeno...

And the plays 🎭 come from @dracor.org of course ♥️.

Using a #GenderClassifier for #Spanish and model-explainability techniques, they identify which textual features most strongly influence the classification of speech as “male” or “female.”
The result? A human-interpretable view of the most gendered elements of dialogue in #Calderón’s #Comedias. 🎭

Keith et al. develop quantitative methods to examine how gender is portrayed across 100+ 17th-century plays by #Calderón.
#CLS #DigitalHumanities #JCLS #LiteraryComputing #Plays #CCLS25 @dracor.org

Instead of trying to normalize this complexity, we embrace it! Our paper offers a statistical-phenomenological look at the complexity of the reading act. Huge thanks to the amazing JCLS editors and reviewers! @jcls-io.bsky.social

As always: #OpenData and #OpenCode
Dekel, Y., Marienberg-Milikowsky, I., & Jacobson, G. A. (2025). "From Readers to Data." #JCLS 2025. Data set. Zenodo. doi.org/10.5281/zeno....
#CCLS2025 #CLS #CitizenScience #Hebrew #LiteraryComputing #CulturalAnalytics

Jacobson et al. explore how #ReaderUncertainty becomes a source of insight by embracing interpretive #ambiguity.
Drawing on 1,026 questionnaire responses from the #HebrewNovelProject, they examine how readers express uncertainty—from skipping questions to outright rejecting interpretive frameworks.

Open Data & Code can be found archived on doi.org/10.5281/zeno...

#DigitalHumanities #CulturalAnalytics #ComputationalLiteraryStudies #CCLS2025 #CLS #JCLS

Furthermore, while #Hades' power correlates negatively with #StoryPopularity, #Persephone's agency correlates positively with it, leading to the fittingly playful title, 'A Powerful Hades Is an Unpopular Dude'.