用情感特征增强嵌入，提升新闻主观性识别准确率。
AI Wizards at CheckThat! 2025: Enhancing Transformer-Based Embeddings with Sentiment for Subjectivity Detection in News Articles
- 在预训练模型中融合情感分数，改进句子表征
- 多语言测试中主观性识别F1显著提升，希腊语达0.51
- 适合关注跨语言文本分析与情感增强的研究者
本文介绍AI Wizards参与CLEF 2025 CheckThat! Lab任务1：新闻文章主观性检测，目标是在单语、多语及零样本设置下对句子进行主观/客观分类。训练与开发数据集涵盖阿拉伯语、德语、英语、意大利语和保加利亚语；最终评估包含希腊语、罗马尼亚语、波兰语、乌克兰语等未见语言，以检验泛化能力。核心方法通过辅助情感模型提取情感分数，并与mDeBERTaV3-base、ModernBERT-base（英语）和Llama3.2-1B的句子表示融合，增强变压器分类器性能，优于标准微调。针对各语言普遍存在的类别不平衡问题，采用开发集优化的决策阈值校准策略。实验表明，情感特征融合显著提升性能，尤其在主观性识别上；该框架在多项指标中排名靠前，其中希腊语达到宏平均F1=0.51，位列第一。
原文摘要 · Abstract (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).
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。