arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2208.04568cs.CLcs.AI
收藏

数据损坏对低资源语言命名实体识别的影响

The Impact of Data Corruption on Named Entity Recognition for Low-resourced Languages

University of the Witwatersrand(威特沃特斯兰德大学)

Manuel Fokam, Michael Beukman

AI总结:

本文系统研究了低资源语言环境下数据数量与质量对预训练语言模型命名实体识别性能的影响,发现完全标注的少量数据优于缺失标签的大量数据,且模型仅用10%训练数据即可表现优异,该结论在多种语言和模型中均成立。

摘要:

数据可用性和质量是低资源语言自然语言处理面临的主要挑战。特别是,其可用数据明显少于高资源语言。这些数据通常质量低下,充斥着错误、无效文本或错误标注。许多先前的工作通过生成合成数据或过滤掉数据集中的低质量部分来解决这些问题。我们转而更深入地研究这些因素,系统地衡量在低资源环境下数据数量和质量对预训练语言模型性能的影响。结果表明,拥有较少的完全标注句子明显优于拥有较多缺失标签的句子;并且模型仅用10%的训练数据就能表现得非常出色。重要的是,这些结果在十种低资源语言、英语和四个预训练模型中保持一致。

英文摘要:

Data availability and quality are major challenges in natural language processing for low-resourced languages. In particular, there is significantly less data available than for higher-resourced languages. This data is also often of low quality, rife with errors, invalid text or incorrect annotations. Many prior works focus on dealing with these problems, either by generating synthetic data, or filtering out low-quality parts of datasets. We instead investigate these factors more deeply, by systematically measuring the effect of data quantity and quality on the performance of pre-trained language models in a low-resourced setting. Our results show that having fewer completely-labelled sentences is significantly better than having more sentences with missing labels; and that models can perform remarkably well with only 10% of the training data. Importantly, these results are consistent across ten low-resource languages, English, and four pre-trained models.

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。

↑