Thanks to our sponsors we are able to offer prizes for the winners of the PolEval competition who will agree to open-source and submit the source code of their methods and the machine learning models used. The authors of the best performing methods in each of the tasks (and subtasks) will be awarded a prize. The winning methods will be chosen according to the rules described on the Tasks page.
- Prizes will be awarded in each of the tasks and subtasks announced for PolEval 2018 competition, i.e.: Task 1: Subtask A, Subtask B, Task 2 and Task 3 (4 prizes in total).
- Prizes will be awarded to the authors of the best performing methods, as evaluated by measuring overall accuracy on the test data set which will be released in due time.
- Prizes will be awarded only to the authors of systems who will agree to submit the source code of the systems used to tag the test data, along with additional information concerning the training data used and the generated machine learning model.
- Prizes will be awarded to individuals or teams of individual researchers, not institutions or companies.
- A single team or individual may qualify for more than one prize, but in this case her/his prize will be reduced by 10% to cover the income tax. In this case we will ask to submit additional personal information to issue the PIT-8C document.
- Do I have to release the source code of my system to participate in PolEval 2018?
No, but you are strongly encouraged to do so. Releasing the source code is a strict condition for receiving the award.
- How many sets of answers can I submit?
One participant (individual or team) may submit up to 3 different sets of answers for each task or subtask. Those answers may come from different versions of your systems, different systems or models, systems with differently fine-tuned parameters, etc.