12th March 2021 - 11am - Jonas Groschwitz: Seminar

TITLE:  AM dependency trees: Making compositional semantic graph parsing work.


In this talk, I will discuss our parser for semantic graphs such as Abstract Meaning Representation (AMR). Our approach combines neural models with mechanisms from compositional semantic construction. At the center of our approach are AM (Apply-Modify) dependency trees, which we developed to both reflect linguistic principles and yield a simple parsing model. In particular, we can find consistent AM dependency trees as the latent compositional structures for our training data, which is crucial when training a compositional parser. The parser then employs neural supertagging and dependency models to predict interpretable, meaningful operations that construct the semantic graph. The result is a semantic parser with strong performance across a range of graphbanks, that also provides insights to the compositional patterns of the graphs. I will also touch on our recent work on normalizing these compositional structures across graphbanks, and on learning them automatically.


Jonas Groschwitz is a Postdoc at Saarland University (Joint PhD 2020 at Saarland, Germany and Macquarie University, Sydney, Australia). His interests lie in semantic parsing and natural language generation, i.e. turning a sentence into meaning and vice versa. His focus lies on models that combine symbolic and neural mechanisms to capture the hidden structure of language.  



Add to your calendar

 vCal  iCal

Mar 12 2021 -

12th March 2021 - 11am - Jonas Groschwitz: Seminar

This event is co-organised by ILCC and by the UKRI Centre for Doctoral Training in Natural Language Processing, https://nlp-cdt.ac.uk

Zoom invitation only