CheckThat! 2025的AI魔法师：融合情感增强Transformer嵌入，用于新闻文章的主观性检测

摘要

本文介绍了AI Wizards团队参与CLEF 2025 CheckThat!实验室任务1：新闻文章中的主观性检测，在单语、多语及零样本场景下对句子进行主观/客观分类的情况。训练/开发数据集涵盖了阿拉伯语、德语、英语、意大利语和保加利亚语；最终评估则引入了未见过的语言（如希腊语、罗马尼亚语、波兰语、乌克兰语）以检验模型的泛化能力。我们的核心策略是通过将辅助模型生成的情感评分与句子表征相结合，来增强基于Transformer的分类器，旨在超越标准的微调方法。我们利用mDeBERTaV3-base、ModernBERT-base（英语）及Llama3.2-1B探索了这一情感增强架构。针对跨语言普遍存在的类别不平衡问题，我们采用了基于开发集优化的决策阈值校准方法。实验结果表明，情感特征的整合显著提升了模型性能，尤其是主观类别的F1分数。这一框架使我们在多个语言上取得了高排名，特别是在希腊语上获得了第一名（宏F1 = 0.51）。

English

This paper presents AI Wizards' participation in the CLEF 2025 CheckThat! Lab Task 1: Subjectivity Detection in News Articles, classifying sentences as subjective/objective in monolingual, multilingual, and zero-shot settings. Training/development datasets were provided for Arabic, German, English, Italian, and Bulgarian; final evaluation included additional unseen languages (e.g., Greek, Romanian, Polish, Ukrainian) to assess generalization. Our primary strategy enhanced transformer-based classifiers by integrating sentiment scores, derived from an auxiliary model, with sentence representations, aiming to improve upon standard fine-tuning. We explored this sentiment-augmented architecture with mDeBERTaV3-base, ModernBERT-base (English), and Llama3.2-1B. To address class imbalance, prevalent across languages, we employed decision threshold calibration optimized on the development set. Our experiments show sentiment feature integration significantly boosts performance, especially subjective F1 score. This framework led to high rankings, notably 1st for Greek (Macro F1 = 0.51).

CheckThat! 2025的AI魔法师：融合情感增强Transformer嵌入，用于新闻文章的主观性检测

AI Wizards at CheckThat! 2025: Enhancing Transformer-Based Embeddings with Sentiment for Subjectivity Detection in News Articles

摘要

Support