Group picture from NLP4Sustain workshop at KONVENS 2025 in Hildesheim. Also a picture of the keynote speaker and the shared task winners.
In September 2025, Jakob Prange and Charlott Jakob (TU Berlin) co-organized a GermEval shared task on analyzing German sustainability reports at KONVENS in Hildesheim. Keynote: Mariana Madruga de Brito (Helmholtz Leipzig). Shared task winners: Julia Wunderle, Jan Pfister, Andreas Hotho!
Marc Schulder is standing in the auditorium of the Faculty of Applied Computer Science at the University of Augsburg.
This week, I had the pleasure of co-hosting Marc Schulder (University of Hamburg) with Elisabeth André at the U' of Augsburg. With his amazing talk on Computational Sign Linguistics, Marc opened our new colloquium series on AI Meets Human Data: Learning from Language, Vision, and Interaction.
Looking forward to #acl2025 in Vienna! Excited to catch up. I am also hiring, one fully paid (100% E13) position for either a PhD student or a PostDoc in computational linguistics / digital humanities / corpus linguistics (earliest start date: October 2025). Who is interested?
Last week, we've been on a retreat to talk about the big picture of our research, research ideas, and synergies between group members – and did some team building.
@annefriedrich.bsky.social
was our external guest and helped us out of a couple of local optima. Thank you for participating!
Caribbean corner beach in Lichtenfels
Just came home from an awesome three-day retreat with the BamNLP group of @romanklinger.de. We talked about uncertainty in NLP, modeling emotions, and knowledge editing. Thank you for making me feel so welcome! I immensely enjoyed hanging out at a Bavarian Caribbean beach with you, BamNLP!
Thanks for the nice feedback. :) My favorite in the subtitles was Chatchipiti. 😆
Picture from inaugural lecture of Prof. Friedrich and Prof. Chiarcos at Uni Augsburg, Germany.
This Wednesday, Christian Chiarcos and I finally gave our inaugural lectures at the University of Augsburg, hosted jointly by the Faculty of Applied Computer Science and the Faculty of Philology and History. Deeply grateful for working in this interdisciplinary context!
Illustration of the components of PEDANTIC: patent text, span annotations for writing issues such as Antecedent Basis or Undefined Term. Binary and Multi-Label Tasks, as well as applying an LLM for pair-wise reason judgments.
Ever wondered how effective LLMs are on detecting problems with formulating patents? Our paper on the new dataset PEDANTIC has just been accepted to PatentSemTech 2025 @ SIGIR 2025 in Padua, Italy. Really proud of Valentin Knappich. Find the prepint here: arxiv.org/abs/2505.21342