Workshop co-located with SLTC-2024
Workshop date: 29 November 2024, before noon
Abstracts due (1 page pdf): 25 October
Notification: 1 November
Universal Dependencies (UD) is an annotation framework for natural language syntax analysed as dependency trees. Its main characteristics are cross-linguality (it has been applied to over 150 languages) and robustness (it can analyse any input, however far it is from standard grammaticality). Several algorithms and tools are available for easy deployment of UD parsing, which makes it into a useful step for pipelines of explainable natural language processing; the UD parse tree itself is an inspectable data structure that allows for both data-driven and rule-based further processing.
The workshop aims to survey existing downstream applications of UD and bring together researchers who want to exchange knowledge and experience about them. Possible topics of presentation include:
- semantic parsing
- language-typological studies
- language education
- text analysis
- information extraction
- tools for UD tree processing
- combinations with other NLP tools
We will not publish proceedings, but accepted abstracts will be made available on the workshop website. Original work is welcome, including work in progress, but we are also interested in previously published work that can complete the picture of what applications of UD there are. Both talks and demos are welcome.
Registration with this form. The deadline is 25 October, and notifications are due 1 November. The registration form asks for a title and a short abstract, in case you want to give a presentation. We will check that your topic is in scope, but try to give room to all presentations that are; it is also possible to participate without a talk.
Organizers
- Arianna Masciolini, University of Gothenburg
- Jyrki Nummenmaa, Tampere University
- Aarne Ranta, Chalmers and University of Gothenburg
- Kirsi Sandberg, Tampere University