Description and task

Two tracks are suggested:

• CoBaLD Parse: the task is to create a parser for the linguistic annotation of English and Russian in the Enhanced CoBaLD standard which includes morphologic, syntactic and semantic levels.
  The standard is based on Enhanced Universal Dependencies, suggests ellipsis restoration in some cases. The semantic pattern is based on the simplified version of the Compreno markup and includes annotation of word meanings and the relations between words. Two datasets are suggested for training — for Russian and English annotated in the given format.
• CoBaLD CLT the task is to create a parser for the linguistic annotation of Hungarian or Serbian languages in the CoBaLD standard with the help of Cross Lingual Transfer.
  Here, the morpho-syntactic markup is based on the basic version of the UD model and does not suggest restoration of the ellipted nodes. As in the previous task, the semantic pattern is based on the simplified version of the Compreno markup. For the task, publish two small datasets are published for Hungarian or Serbian provided with human-checked annotation.

Timeline

01.06.2024 — final dataset publication for CoBaLD Parse track;

15.06.2024 — dataset publication for CoBaLD CLT track;

20.08.2024 — Shared Task ends, results;

28.08.2024 — deadline for paper submissions.