DeMeVa在LeWiDi-2025：基于上下文学习与标签分布学习的多视角建模

DeMeVa at LeWiDi-2025: Modeling Perspectives with In-Context Learning and Label Distribution Learning

September 11, 2025

作者: Daniil Ignatev, Nan Li, Hugh Mee Wong, Anh Dang, Shane Kaszefski Yaschuk

cs.AI

摘要

本系统论文介绍了DeMeVa团队在第三届“学习分歧”共享任务（LeWiDi 2025；Leonardelli等人，2025）中的研究方法。我们探索了两个方向：一是基于大语言模型的上下文学习（ICL），在此我们对比了不同的示例采样策略；二是结合RoBERTa（Liu等人，2019b）的标签分布学习（LDL）方法，我们评估了多种微调技术。我们的贡献主要体现在两个方面：（1）我们证明了ICL能够有效预测特定标注者的注释（即视角主义注释），并且将这些预测聚合为软标签后，能取得具有竞争力的性能；（2）我们提出LDL方法在软标签预测方面展现出潜力，值得视角主义研究社区进一步深入探索。

English

This system paper presents the DeMeVa team's approaches to the third edition of the Learning with Disagreements shared task (LeWiDi 2025; Leonardelli et al., 2025). We explore two directions: in-context learning (ICL) with large language models, where we compare example sampling strategies; and label distribution learning (LDL) methods with RoBERTa (Liu et al., 2019b), where we evaluate several fine-tuning methods. Our contributions are twofold: (1) we show that ICL can effectively predict annotator-specific annotations (perspectivist annotations), and that aggregating these predictions into soft labels yields competitive performance; and (2) we argue that LDL methods are promising for soft label predictions and merit further exploration by the perspectivist community.