SubmissionNumber#=%=#6
FinalPaperTitle#=%=#Fine-Tuning Language Models on Dutch Protest Event Tweets
ShortPaperTitle#=%=#
NumberOfPages#=%=#
CopyrightSigned#=%=#
JobTitle#==#
Organization#==#
Abstract#==#Being able to obtain timely information about an event, like a protest, becomes increasingly more relevant with the rise of affective polarisation and social unrest over the world. Nowadays, large-scale protests tend to be organised and broadcast through social media. Analysing social media platforms like X has proven to be an effective method to follow events during a protest. Thus, we trained several language models on Dutch tweets to analyse their ability to classify if a tweet expresses discontent, considering these tweets may contain practical information about a protest. Our results show that models pre-trained on Twitter data, including Bernice and TwHIN-BERT, outperform models that are not. Additionally, the results showed that Sentence Transformers is a promising model. The added value of oversampling is greater for models that were not trained on Twitter data. In line with previous work, pre-processing the data did not help a transformer language model to make better predictions.
Author{1}{Firstname}#=%=#Meagan
Author{1}{Lastname}#=%=#Loerakker
Author{1}{Username}#=%=#meagan
Author{1}{Email}#=%=#meaganloerakker@gmail.com
Author{1}{Affiliation}#=%=#Netherlands Police
Author{2}{Firstname}#=%=#Laurens
Author{2}{Lastname}#=%=#Müter
Author{2}{Username}#=%=#laurens.muter
Author{2}{Email}#=%=#l.h.f.muter@uu.nl
Author{2}{Affiliation}#=%=#Netherlands Police, Utrecht, the Netherlands
Author{3}{Firstname}#=%=#Marijn
Author{3}{Lastname}#=%=#Schraagen
Author{3}{Username}#=%=#marijnschraagen
Author{3}{Email}#=%=#m.p.schraagen@uu.nl
Author{3}{Affiliation}#=%=#Utrecht University

==========
èéáğö