Wessel Poelman

I work at the KU Leuven as a PhD student in the LAGoM group of Miryam de Lhoneux.

I’m interested in multilingual natural language processing (NLP) in a broad sense (from tokenization to evaluation). My current focus is on experimental practices in multilingual NLP; how do we know we’re making the progress we think we’re making? This includes carefully thinking about which languages are used in experiments, which assumptions are made about model architectures and evaluations, and how differences between languages impact this.

Before coming to Belgium, I worked at the Technical University of Munich in Germany for half a year as a research associate (Wissenschaftlicher Mitarbeiter). Before that, I completed a master’s in Information Science at the University of Groningen and worked for about two years as a machine learning engineer at Web-IQ. You can find more details on my about page.

Feel free to reach out:

wessel.poelman at kuleuven.be

or

contact “apenstaartje” wesselpoelman.nl

| KU Leuven | Github | LinkedIn | Semantic Scholar | Google Scholar |

Non-academic writing

I occasionally write about (strange) things I find interesting. I also maintain some lists, which you can find at the top.

Projects

  1. qwanqwa: a toolkit to work with a large variety of language metadata (identifiers, speaker information, locales, …)
  2. planit: declarative SLURM workflows in Python
  3. typdiv-sampling: a framework to create and evaluate language samples (together with Esther Ploeger)

Selected Publications

News