Natural Language Learning and Generation Lab

Prof. Dr. Steffen Eger

NLLG focuses on problems of text generation (also in humanities contexts; such as poetry generation), evaluation of text generation, as well as problems at the intersection of Natural Language Processing (NLP), the digital humanities and social science. Sample problems include:

  1. How to robustly, efficiently, etc. evaluate machine translation systems?
  2. How to trace (anti-)solidarity with refugees change over time?
  3. How to support scientists in automatic figure generation?
  4. How to induce high-quality automatic poetry generation systems?
  5. How well does literary translation with Large Language Models (LLMs) work?

Prof. Dr. Steffen Eger
Professor of Natural Language Learning and Generation

Current Research Projects

“Metrics4NLG” (2022-2025)

BMBF funded research group including 3 PhD students on the topic of robust, efficient, explainable and high-quality evaluation metrics for text generation (machine translation and summarization), also in humanities contexts.

“Cross-lingual Cross-temporal Evaluation in NLP and the Digital Humanities” (2022-2027)

DFG Heisenberg programme. Funds the interdisciplinary research activities of Steffen Eger until 2027.

“EMCONA” (2024-2027)

DFG funded project on the interplay between emotions and convincingess in argumentation, including emotion-bias free automatic argument generation. Joint with Roman Klinger (University of Bamberg).

Completed Research Projects

Eval4NLP workshop funding (2021)

AI Journal funding for explainable evaluation metrics shared task.

Doubly annotated Corpora (2020-2021)

FiF funded project on metaphor annotation in historical German corpora.

Publications

List of publications of Prof. Dr. Steffen Eger on Google Scholar

List of publications of Prof. Dr. Steffen Eger on dblp

You still have questions?

Then contact our Natural Language Understanding Lab.

 nllg@utn.de