Dear colleagues,
We are excited to announce the third edition of the Shared Task on Multilingual Coreference Resolution! The focus of the task is learning models for prediction of coreference across a range of typologically different languages. Both identification of mentions and clustering of mentions that refer to the same entity are included. This year, we make a special focus on identification of zeros (e.g. zero subjects, objects, ellipses). However, we don't mean it as a limitation. You can participate just for selected languages and without dealing with zero mentions.
The Shared Task is based on the CorefUD 1.2. dataset, which deals with coreference resolution in 15 languages.
EXPECTED TIMELINE
already started - development phase
gold training and development data available at the LINDAT/CLARIAH-CZ repository
UDPipe-reparsed development data available
evaluation tool available on GitHub
baseline system available at GitHub
development evaluation possible via CodaLab
June 3, 2024 - start of the evaluation phase: test evaluation via CodaLab
June 17, 2024 (AoE) - end of evaluation phase
August, 2024 - submission of system description papers to the CRAC 2024 workshop
between November 12 and 16, 2024 – the CRAC workshop at EMNLP
If you consider participation, please let us know by filling in the registration form as soon as possible:
https://ufal.mff.cuni.cz/corefud/crac24#registration
You are very welcome to participate!
Anna Nedoluzhko
on behalf of the shared task organizers