The CLEF-2024 CheckThat! Lab: check-worthiness, subjectivity, persuasion, roles, authorities, and adversarial robustness.

Barrón Cedeño, Alberto; Alam, Firoj; Chakraborty, Tanmoy; Elsayed, Tamer; Nakov, Preslav; Przybyła, Piotr; Struß, Julia Maria; Haouari, Fatima; Hasanain, Maram; Ruggeri, Federico; Song, Xingyi; Suwaileh, Reem

doi:http://dx.doi.org/10.1007/978-3-031-56069-9_62

The CLEF-2024 CheckThat! Lab: check-worthiness, subjectivity, persuasion, roles, authorities, and adversarial robustness.

Citación

Barrón-Cedeño A, Alam F, Chakraborty T, Elsayed T, Nakov P, Przybyła P, et al. The CLEF-2024 CheckThat! Lab: check-worthiness, subjectivity, persuasion, roles, authorities, and adversarial robustness. In: Goharian N, Tonellotto N, He Y, Lipani A, McDonald G, Macdonald C, Ounis I, editors. Advances in Information Retrieval. 46th European Conference on Information Retrieval, ECIR 2024; 2024 March 24–28; Glasgow, UK. Cham: Springer; 2024. p. 449–58. DOI: 10.1007/978-3-031-56069-9_62

Resumen

The first five editions of the CheckThat! lab focused on the main tasks of the information verification pipeline: check-worthiness, evidence retrieval and pairing, and verification. Since the 2023 edition, it has been focusing on new problems that can support the research and decision making during the verification process. In this new edition, we focus on new problems and —for the first time— we propose six tasks in fifteen languages (Arabic, Bulgarian, English, Dutch, French, Georgian, German, Greek, Italian, Polish, Portuguese, Russian, Slovene, Spanish, and code-mixed Hindi-English): Task 1 estimation of check-worthiness (the only task that has been present in all CheckThat! editions), Task 2 identification of subjectivity (a follow up of CheckThat! 2023 edition), Task 3 identification of persuasion (a follow up of SemEval 2023), Task 4 detection of hero, villain, and victim from memes (a follow up of CONSTRAINT 2022), Task 5 Rumor Verification using Evidence from Authorities (a first), and Task 6 robustness of credibility assessment with adversarial examples (a first). These tasks represent challenging classification and retrieval problems at the document and at the span level, including multilingual and multimodal settings.