The aim of the competition is to determine whether a claim in a tweet is worth fact-checking. Task 1 asks to detect relevant tweets on the basis of different complementary criteria: check-worthiness, verifiability, harmfulness, and attention-worthiness. Manually annotated data are provided for these four subtasks in five languages: Arabic, Bulgarian, Dutch, English, and Turkish. For a sixth language, Spanish, a larger-scale dataset annotated by investigative journalists was provided, but only for the subtask of check-worthiness identification.
Forum
Year
2022
Link to publication
