The training corpus comprises 4,828 samples from 1,320 reviews about the accommodation sector.

The training data come from the ABSAPT22 plus newly annotated data.

The test corpus comprises 283 unique review samples (task 1) and 1,176 samples from 282 reviews (task 2).

All test data is new.

🆕🆕🆕LINK to Dataset.🆕🆕🆕

*Note: each review can hold opinions about several aspects.

Participants will receive the compressed corpus and the password in the subscription email.

Register here

absapt@inf.ufpel.edu.br.

--------------------

After the competition you can ask for the dataset access through this form.