Sign in


Semantic Role labeling for Dutch. Understanding who does what, where, to whom.


What Stroll can do for you

  • Fast and high accuracy
  • Easy to use: integrates with Stanza, formerly StanfordNLP, and the VU Readingmachine

Stroll is trained on the largest publically available dataset for Dutch, SoNaR.
For the given input sentences, it finds the frames (typically the verbs), and its corresponding arguments.
There are 4 arguments, Arg0 - Arg4, and 13 modifiers, ArgM-*.

Programming language
  • Python 100%
  • Apache-2.0
</>Source code

Participating organisations

Social Sciences & Humanities
Netherlands eScience Center
Vrije Universiteit Amsterdam


Contact person

Jisk Attema

Jisk Attema

Netherlands eScience Center
Mail Jisk
Dafne van Kuppevelt
Dafne van Kuppevelt
Netherlands eScience Center
Jisk Attema
Jisk Attema
Netherlands eScience Center

Related projects

Inside the filter bubble

A framework for deep semantic analysis of mobile news consumption traces

Updated 3 months ago