[Corpora-List] DISRPT2021 - discourse relation task - test data release and surprise language

Amir Zeldes Amir.Zeldes at georgetown.edu
Mon Jul 5 11:43:57 CEST 2021


Dear colleagues,

We are happy to announce the release of test data for this year’s DISRPT shared task �on discourse relations across formalisms, as well as a surprise language dataset: the Persian RST Corpus (https://github.com/hadiveisi/PersianRST). Training and development data can now be found at the repository for the shared task, which is co-located with CODI 2021 at EMNLP 2021:

https://github.com/disrpt/sharedtask2021

The repository covers three tasks, using a uniform format across RST, SDRT and PDTB for a variety of languages and genres:

* Task 1 - �Discourse Segmentation, * Task 2 - �Connective Identification, * Task 3 - �Relation Classification.

Contact email: � <mailto:disrpt2021 at googlegroups.com> disrpt2021 at googlegroups.com Official website: � <https://sites.google.com/georgetown.edu/disrpt2021> https://sites.google.com/georgetown.edu/disrpt2021 ​​​​​ Tentative schedule:

* March, 2021 - shared task sample data release * May, �2021 - training data release * July, 2021 - test data release * August, 2021 - papers and systems due * September, 2021 - notification of acceptance * October, 2021 - camera-ready papers due * November, 2021 - CODI workshop

All the best,

Organizers: Amir Zeldes, Mikel Iruskieta, Janet Liu, Philippe Muller and Chloé Braud

-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/html Size: 9100 bytes Desc: not available URL: <https://mailman.uib.no/public/corpora/attachments/20210705/f513d46e/attachment.txt>



More information about the Corpora mailing list