Bayesian approaches for preference-based response generation in dialogue (BAPAGiD)

This project investigates how game theoretical methods can be applied to the generation of pragmatically optimal answers that contain both numerals and adjectives as well as to the generation of so-called speech-act conditionals. Each of these linguistic means can trigger implicatures concerning preferences of the dialogue participants. These implicatures will be modelled by Bayesian probabilistic approaches and subsequently empirically evaluated.

Publications

  • Hesse, Christoph & Anton Benz. 2020. Scalar bounds and expected values of comparatively modified numerals. Journal of Memory & Language, Vol. 111.
  • Hesse, Christoph & Anton Benz. 2018. Giving the wrong impression: Strategic use of comparatively modified numerals in a question answering system. Proceedings of KONVENS 2018. Vienna, Austria.
  • M. Mohammadi, J. Fischer, M. Langner & R. Klabunde. 2018. Conveying the user’s intention by generating speech act conditionals as indirect answers. Proceedings of KONVENS 2018, 80-88.
  • Mohammadi, Maryam, Christoph Hesse, Maurice Langner, Judith Fischer, Anton Benz & Ralf Klabunde. Communicating an understanding of intention: Speech act conditionals and modified numerals in a Q/A system. Proceedings of the Workshop on the SemDial 22. Aix-en-Provence, France.