融合多语言特征与上下文信息，预测不同母语者词汇难度。
UOL@IDEM at BEA 2026 Shared Task 1: Neural Fusion and Feature-Rich Modeling for L1-Aware Vocabulary Difficulty Prediction
- 用上下文嵌入+人工设计特征联合建模词汇难度
- 中文、德语、西班牙语预测误差分别低至0.891、1.037、1.132
- 频率和语义对齐等特征能有效捕捉母语差异
本文介绍乌尔斯特大学@IDEM团队在BEA 2026共享任务中关于母语感知词汇难度预测的闭源提交。我们将任务建模为回归问题，分别为西班牙语、德语和中文（以下简称中文）训练独立系统。系统结合多语言上下文表示与人工特征，包括频率、表层形式、检索证据、语义对齐、词源相似性和掩码语言模型可预测性。开发结果表明，相比官方基线有持续提升，其中句子嵌入编码器如BGE-M3、multilingual E5和LaBSE表现最佳。正式提交的测试集RMSE分别为：西班牙语1.132，德语1.037，中文0.891。特征分析显示频率是最稳定的预测因子，而上下文可预测性、形式相似性、检索信号和语义特征提供了互补的母语敏感信号。错误分析发现模型排名效果好，但对最简单词汇存在过预测问题。
原文摘要 · Abstract (English)
This paper describes UOL@IDEM's closed-track submission to the BEA 2026 shared task on L1-aware vocabulary difficulty prediction. We model the task as regression and train separate systems for Spanish, German, and Mandarin Chinese\footnote{Below we use \emph{Chinese} for brevity.}. Our system combines multilingual contextual representations with engineered features capturing frequency, surface form, retrieval evidence, semantic alignment, cognate similarity, and masked-language-model predictability. Development results show consistent gains over the official closed-track baselines, with sentence-embedding encoders such as BGE-M3, multilingual E5, and LaBSE performing best. Official submissions achieve RMSE scores of 1.132, 1.037, and 0.891 for Spanish, German, and Chinese, respectively. Feature analysis identifies frequency as the most stable predictor, while contextual predictability, form similarity, retrieval, and semantic features provide complementary L1-sensitive signals. Error analysis shows strong ranking performance but weaker calibration for the easiest items, which are often overpredicted. See https://github.com/Nouran-Khallaf/UoL-IDEM-BEA2026-Vocabulary-Difficulty-Prediction
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。