Communication Dans Un Congrès Année : 2025

ClimateSense at CheckThat! 2025: Combining Fine-tuned Large Language Models and Conventional Machine Learning Models for Subjectivity and Scientific Web Discourse Analysis

Pasquale Lisena
Enrico Daga
  • Fonction : Auteur
  • PersonId : 1560231
Raphael Troncy
Harith Alani
  • Fonction : Auteur
  • PersonId : 1434532

Résumé

These working notes present the ClimateSense team participation in the CheckThat! 2025 Lab Challenge for tasks 1 and 4a that investigated: 1) the subjectivity of news article sentences, and; 2) the detection of scientific content in social media posts. Pre-trained Large Language Models (LLMs), conventional Machine Learning (ML) models, sentence encoders, data augmentation, and filtering techniques were leveraged by the ClimateSense team to investigate these tasks. In this paper, we detail the approaches for each task, present the methodology, and report on the performance of each submission. The fine-tuning of pre-trained models shows particularly strong results for Task 4a, where we achieved the first rank on the final evaluation leaderboard. This result shows that LLMs can benefit from lightweight traditional classification models when performing classification tasks.

Fichier principal
Vignette du fichier
CLEF_2025___CheckThat.pdf (648 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05150274 , version 1 (08-07-2025)

Licence

Identifiants

  • HAL Id : hal-05150274 , version 1

Citer

Grégoire Burel, Pasquale Lisena, Enrico Daga, Raphael Troncy, Harith Alani. ClimateSense at CheckThat! 2025: Combining Fine-tuned Large Language Models and Conventional Machine Learning Models for Subjectivity and Scientific Web Discourse Analysis. CLEF 2025 Working Notes, Sep 2025, Madrid, Spain. ⟨hal-05150274⟩

Collections

442 Consultations
236 Téléchargements

Partager

  • More