We are TULP


  • Affective Language Production: Content selection, message formulation and computational modelling (ALP, NWO Free Competition)
    link: alp.uvt.nl
    In this project we study how speaker’s emotional state influences the language that they produce, looking both at the early stages (content selection) and later stages of language production (message formulation). In addition, we develop a computational model that generates emotionally charged texts, paving the way for targeted news-reporting.
  • Second language tutoring using social robots (L2TOR, H2020)
    link: www.l2tor.eu
    L2TOR (pronounced ‘el tutor’) is a scientific research project funded by the Horizon 2020 programme of the European Commission. The project aims to design a child-friendly tutor robot that can be used to support teaching preschool children a second language (L2), by interacting with them in their own social and referential world.
  • The Automated Newsroom (RAAK-SIA)
    link: futuremedialab.nl/onderzoekrobotjournalistiek/automatische-nieuwsredactie/
    This project on the one hand aims to develop new tools that can support journalists, including bots that automatically generate reports. On the other hand, it also studies what the impact is of such new technologies on how professionals and the public perceive such changes in the newsroom.
  • Automatic summarization of discussion forum threads (DISCOSUMO, NWO Creative Industry)
    link: discosumo.ruhosting.nl
    People access the internet increasingly through mobile devices. In this project, we study how online discussions can be automatically summarised, so that they can be read easier on mobile devices.
  • Understanding cancer treatment data: Using data science to help cancer patients during treatment decision making (DSC/t)
    After a cancer diagnosis, oncologists are obliged to inform patients about the chances of a favourable effect (e.g., long-term survival) and the risks of adverse effects (e.g., death, side-effects) of treatment options. The aim of this project is to analyse the data of millions of Dutch cancer patients, determining what the pros and cons of different treatment options are for individual patients, and automatically presenting individualised predictions, with the aim of facilitating shared decision making.

Older projects:

  • Bridging the gap between psycholinguistics and computational linguistics: The case of referring expressions (NWO VICI)
    link: bridging.uvt.nl
    In this project, we studied the production of referring expressions (phrases that refer to entities in the world, including descriptions such as “the man in a suit” or “the younger-looking man”) combining insights from psycholinguistics and computational modelling.