@inproceedings{hanif-etal-2025-university,
    title = "{U}niversity of {I}ndonesia at {S}em{E}val-2025 Task 11: Evaluating State-of-the-Art Encoders for Multi-Label Emotion Detection",
    author = "Hanif, Ikhlasul  and
      Yulianrifat, Eryawan Presma  and
      Ongris, Jaycent  and
      Tjitrahardja, Eduardus  and
      Azmi, Muhammad  and
      Naufal, Rahmat  and
      Wicaksono, Alfan",
    editor = "Rosenthal, Sara  and
      Ros{\'a}, Aiala  and
      Ghosh, Debanjan  and
      Zampieri, Marcos",
    booktitle = "Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/transition-to-people-yaml/2025.semeval-1.279/",
    pages = "2149--2164",
    ISBN = "979-8-89176-273-2",
    abstract = "This paper presents our approach for SemEval 2025 Task 11 Track A, focusing on multilabel emotion classification across 28 languages. We explore two main strategies: fully fine-tuning transformer models and classifier-only training, evaluating different settings such as fine-tuning strategies, model architectures, loss functions, encoders, and classifiers. Our findings suggest that training a classifier on top of prompt-based encoders such as mE5 and BGE yields significantly better results than fully fine-tuning XLMR and mBERT. Our best-performing model on the final leaderboard is an ensemble combining multiple BGE models, where CatBoost serves as the classifier, with different configurations. This ensemble achieves an average F1-macro score of 56.58 across all languages."
}