The speaker speaks on site.
Abstract:
This presentation explores the use of computational methods to identify characters‘ emotions in dramatic texts. It begins with a humanities-based approach to conceptualizing emotions, followed by ontology-based transformer model training. The process involves close reading for annotation and the subsequent classification of emotions through automated methods. The presentation also examines the interplay between historical-hermeneutic approaches to understanding emotions and quantitative methods that analyze patterns and trends. A key focus is the comparison between qualitative interpretations and the processing of large-scale text data using algorithms. Specifically, the emotions “anger” and “being moved” are analyzed, focusing on their linguistic representation and emotional trajectories in dramatic contexts. The aim is to highlight how the combination of humanities research and computational techniques can provide new insights into the development and depiction of emotions in literary works.
Short bio:
Katrin Dennerlein is a private lecturer at the University of Würzburg. In 2024, she was a Max Kade Professor at the University of Illinois at Chicago for one semester. From 2020 to 2023, she led the DFG project "Emotions in Drama" and also had her own position in it. She completed her doctorate on the narratology of space and her habilitation on the history of German-language comedy from the 17th to the early 19th century. Her research focuses on drama and theater, the novel of the Goethe era, space and mobility in media narration, and digital humanities.
Her most important publications include Narratologie des Raumes (2009), Materialien und Medien der Komödienschgeschichte (2021). Recent papers have been written on the first known version of Beauty and the Beast and on digital research into emotions in 18th century drama subgenres and emotions in director's notes from this period.