arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2301.09685cs.CLcs.LG
收藏

噪声并行数据对齐

Noisy Parallel Data Alignment

George Mason University(乔治梅森大学)

Ruoyu Xie, Antonios Anastasopoulos

摘要:

当前自然语言处理的一个持续挑战是,其主要进展如何倾向于不成比例地支持资源丰富的语言,而将大量资源不足的语言抛在后面。由于缺乏训练和评估模型所需的资源,大多数现代语言技术要么不存在,要么不可靠,无法处理濒危的、本地的和非标准化的语言。光学字符识别(OCR)常用于将濒危语言文档转换为机器可读数据。然而,这样的OCR输出通常是有噪声的,并且大多数单词对齐模型不是为了在这样的噪声条件下工作而构建的。在这项工作中,我们研究了现有的词级对齐模型在嘈杂的设置,旨在使他们更强大的噪音数据。我们的噪声模拟和结构偏置方法,在多个语言对上进行测试,设法将最先进的基于神经的对齐模型的对齐出错率降低到59.6%。

英文摘要:

An ongoing challenge in current natural language processing is how its major advancements tend to disproportionately favor resource-rich languages, leaving a significant number of under-resourced languages behind. Due to the lack of resources required to train and evaluate models, most modern language technologies are either nonexistent or unreliable to process endangered, local, and non-standardized languages. Optical character recognition (OCR) is often used to convert endangered language documents into machine-readable data. However, such OCR output is typically noisy, and most word alignment models are not built to work under such noisy conditions. In this work, we study the existing word-level alignment models under noisy settings and aim to make them more robust to noisy data. Our noise simulation and structural biasing method, tested on multiple language pairs, manages to reduce the alignment error rate on a state-of-the-art neural-based alignment model up to 59.6%.

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。

↑