数据堂—472万组汉维平行语料数据

汉维平行语料数据包括472万组汉语、维语平行互译语料，数据存储格式为txt文档。已进行数据清洗脱敏质检，可作为文本类数据分析的基础语料库，用于机器翻译等领域

This Chinese-Uyghur parallel corpus consists of 4.72 million pairs of mutually translated Chinese and Uyghur text segments. All data is stored in TXT format, and has been subjected to data cleaning, de-identification and quality inspection. It can serve as a foundational corpus for text data analysis and be applied in fields such as machine translation and other relevant domains.

搜集汇总

背景与挑战

背景概述

该数据集包含472万组汉维平行语料，专用于机器翻译测试任务。数据经过清洗和质量检验，以txt格式存储，平均中文句长为22字符，准确率约为90%。