汉传佛典汉英平行语料库建设研究

On Constructing a Chinese-English Parallel Corpus of Chinese Buddhist Scriptures

汉传佛典在英语世界的译介是中华传统文化外译的重要组成部分,汉传佛典汉英平行语料库的创建有助于中华传统文化的翻译与传播。基于汉传佛典汉英平行语料库建设的背景和现状,运用语料库语言学、机器翻译理论和中古汉传佛典语言的相关特点,对汉传佛典汉英平行语料库建设过程中的语料采集、语料对齐、文本分词等进行了分析,并对翻译记忆库、术语库和语料库的相关应用进行了实操性较强的阐释,从而创建一个汉英字数达到1000万级别的大型汉传佛典汉英平行语料库,为神经网络机器翻译模型预训练提供精确的双语语料,提升机器翻译佛典的质量,以促进中华文化典籍外译模式的推陈出新,助力中华传统文化“走出去”。

The translation of Chinese Buddhist scriptures into English is an important part of translating traditional Chinese culture into the world.The construction of a Chinese-English parallel corpus of Chinese Buddhist scriptures will facilitate the translation and dissemination of traditional Chinese culture.Based on the background and current situation of the construction of the Chinese-English parallel corpus of Chinese Buddhist scriptures,the corpus linguistics,machine translation theory and relevant features of the language of the ancient Chinese Buddhist scriptures are applied to analyze the text collection,alignment and segmentation in the process of the construction of this parallel corpus of Chinese Buddhist scriptures.And the related applications of translation memory,terminology and corpus are explained in a practical way.It aims to build a large Chinese-English parallel corpus of Chinese Buddhist scriptures with 10 million Chinese and English words so as to provide accurate bilingual corpus for the pre-training of neural network machine translation models and to improve the quality of machine translation of Buddhist scriptures.This will promote the innovation of translation mode of Chinese cultural classics and going abroad of traditional Chinese culture.