SubmissionNumber#=%=#232
FinalPaperTitle#=%=#whatdoyoumeme at SemEval-2024 Task 4: Hierarchical-label aware cross-lingual persuasion detection using translated texts
ShortPaperTitle#=%=#
NumberOfPages#=%=#7
CopyrightSigned#=%=#Boshko Koloski
JobTitle#==#
Organization#==#Jozef Stefan Institute
Abstract#==#In this paper, we detail the methodology of team whatdoyoumeme for the SemEval 2024 Task on Multilingual Persuasion Detection in Memes. We employ a cross-lingual approach, utilizing translation to adapt model to Macedonian, Arabic, and Bulgarian, and integrate hierarchical label information to refine detection capabilities. Our methodology encompasses both the analysis of meme content and the application of hierarchical transformations to labels. The effectiveness of the approach is demonstrated through improved model performance in multilingual contexts, highlighting the utility of translation-based methods and hierarchy aware learning, over traditional baselines.
Author{1}{Firstname}#=%=#Nishan
Author{1}{Lastname}#=%=#Chatterjee
Author{1}{Username}#=%=#nishanchatterjee
Author{1}{Email}#=%=#inishan.chatterjee@gmail.com
Author{1}{Affiliation}#=%=#Institut Jožef Stefan, University of La Rochelle
Author{2}{Firstname}#=%=#Marko
Author{2}{Lastname}#=%=#Pranjic
Author{2}{Email}#=%=#marko.pranjic@ijs.si
Author{2}{Affiliation}#=%=#Jozef Stefan Institute
Author{3}{Firstname}#=%=#Boshko
Author{3}{Lastname}#=%=#Koloski
Author{3}{Username}#=%=#boshkok
Author{3}{Email}#=%=#boshko.koloski@ijs.si
Author{3}{Affiliation}#=%=#Jozef Stefan Institute
Author{4}{Firstname}#=%=#Lidia
Author{4}{Lastname}#=%=#Pivovarova
Author{4}{Username}#=%=#lidia.pivovarova
Author{4}{Email}#=%=#lidia.pivovarova@gmail.com
Author{4}{Affiliation}#=%=#University of Helsinki
Author{5}{Firstname}#=%=#Senja
Author{5}{Lastname}#=%=#Pollak
Author{5}{Username}#=%=#spollak
Author{5}{Email}#=%=#senja.pollak@ijs.si
Author{5}{Affiliation}#=%=#Jožef Stefan Institute

==========
èéáğö