Kai Krüger

Computational Linguist at BIBB. I blog about all things NLP, AI and Cognitive Linguistics.

Contact Details

  • For my research, you can contact me at kai (dot) krueger (at) bibb (dot) de
  • If you want to contact me for any matter concerning my teaching at the University of Koblenz, please use kkrueger (at) uni-koblenz (dot) de
  • You can also always shoot me a message at the social media accounts linked here on this website. I am always glad to help, if I can!

Bio

  • February 2022: Research Associate in Computational Linguistics at the German Federal Institute for Vocational Education and Training (BIBB).
  • M.A., Linguistics and Digital Humanities, University of Paderborn (2022)
  • B.A., Linguistics, University at Paderborn (2019)

Publications, Talks, Teaching, etc.

coming soon

My Research Philosophy

My current research focus within the area of Natural Language Processing (NLP) lies in processing online job advertisements (ojas) to develop them as a rich data source for vocational education and training research. To find out more about our specific project, visit RADAR.
While I do very much belong to the large group of people, who are exited by deep learning, transformers and large pretrained language models and do also use them in my research, I am also a strong advocate for using these techniques in a scientific way, which means considering the big picture. This includes conducting epistemologically valuable research by defining research questions and critically reflecting the appropiateness of conclusions on the basis of what in the end are statistical patterns for the real world realities we seek to study. It also includes semantically modelling categories or concepts I am defining on the basis of a robust theoretical foundament. Lastly, it also means to review, reflect, collect and optimize new and old data, to make NLP methods more universally applicable.
All of this is especially true, when NLP is used, like in my research, to facilitate research from other disciplines. These research disciplines come with their own wants and needs. The points mentioned above often heavily influence the results, which, in my experience, is commonly underestimated by researchers not coming from NLP / Computer Science. Researchers from the latter on the other hand tend to undererstimate the influence of sublte differences in their methodology within the context of other disciplines. To bridge this gap, at least for those research areas that put everything human at the core of their interest, is the job of the Digital Humanitists, which I most certainly see myself as.
Therefore, NLP, to me, is more than the F1 scores I am trying to beat. It is about using the force of large scale pattern recognition and generalization algorithms powered by machine learning (elsewhere commonly reffered to as Artificial Intelligence) to enable research, save valuable time and resources and hopefully help people by making their lives easier, happier and more exciting. Or in other words, to take tiny steps in making the world a better place, which to me should always be the ultimate goal of any research. It is just one that is sometimes easy to to lose sight of in the heat of technological advancement.
Besides my current research, I am also passionate about cognitive linguistics. Specifically, I am interested in construction grammar, conceptual bleding, metaphors and frame semantics. I strongly believe these concepts capture major aspects of language and our mind very well and deserve greater recognition. You will frequently find these topics incorporated into my blog posts.