通过对齐稀疏上下文词表示来应对跨语言词义消歧中的多语言诅咒
Combating the Curse of Multilinguality in Cross-Lingual WSD by Aligning Sparse Contextualized Word Representations
AI总结：
本文提出结合预训练单语模型与上下文映射，利用稀疏上下文表示，在跨语言零样本WSD中平均F分数提升6.5点（62.0至68.5），并公开代码。
摘要：
在本文中，我们主张在跨语言零样本词义消歧（WSD）中使用大型预训练单语语言模型，并结合上下文映射机制。我们还报告了严格的实验，证明了通过字典学习过程获得的稀疏上下文词表示的有效性。我们的实验结果表明，上述修改在17种类型多样的目标语言集合上，平均F分数显著提高了近6.5个百分点（从62.0提高到68.5）。我们在https://github.com/begab/sparsity_makes_sense发布了源代码，以便复现我们的实验。
英文摘要：
In this paper, we advocate for using large pre-trained monolingual language models in cross lingual zero-shot word sense disambiguation (WSD) coupled with a contextualized mapping mechanism. We also report rigorous experiments that illustrate the effectiveness of employing sparse contextualized word representations obtained via a dictionary learning procedure. Our experimental results demonstrate that the above modifications yield a significant improvement of nearly 6.5 points of increase in the average F-score (from 62.0 to 68.5) over a collection of 17 typologically diverse set of target languages. We release our source code for replicating our experiments at https://github.com/begab/sparsity_makes_sense.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
