POLEVAL 2020

Welcome to PolEval 2020!

PolEval is a SemEval-inspired evaluation campaign for natural language processing tools for Polish. Submitted tools compete against one another within certain tasks selected by organizers, using available data and are evaluated according to pre-established procedures.

In 2020 we have announced the following tasks:

1. Post-editing and rescoring of automatic speech recognition results
2. Morphosyntactic tagging of Middle, New and Modern Polish
3. Word sense disambiguation
4. Information extraction and entity typing from long documents with complex layouts

PolEval introduction (in Polish)

Info

September 26, 2020

• Thank you for joining in for the PolEval session at the AI & NLP conference! Most presentations and the workshop proceedings are already available.

September 15, 2020

• PolEval 2020 results have been published. Congratulations to all winners!

September 11, 2020

• Thank you for taking part in PolEval tasks! This year we received a record number of over 40 submissions. The results will be announced very soon.

August 21, 2020

• Submission deadline for all tasks has been extended until September 4.

August 6, 2020

• Test data for Task 1, Task 2, Task 3
  and Task 4 has been uploaded.

January 24, 2020

• Description of tasks has been updated.