面向机器翻译的汉英语句格式转换研究

An MT-oriented Study of Chinese-English Sentence Format Transfer

自然语言的语句形式纷繁复杂、变化多端,如何采用一种有效的办法对其进行分类并形式化,一直是困扰计算语言学界的一个难题。HNC理论有关语义块、句类以及语句格式的思想,以及对现代汉语语句的各种基本格式及其变化形式进行的具有确定性和穷尽性的编码处理,使计算机能够透过语言复杂多变的表象,达到以“有限驾驭无限”的目的。本文作为基于HNC理论的汉英机器翻译引擎研究的一部分,旨在运用HNC语句格式理论探索汉英句式转换的一般性规律。

The categorization and formalization of sentences in any natural language has always been a challenge to computational linguistics. The ideas of semantic chunk, sentence category （SC） and sentence format （SF） enable the HNC （Hierarchical Network of Concepts） theory to encode all the basic sentences and their derived ones in modern Chinese in a definite and complete way. As part of the on-going research on HNC-based Chinese-English machine translation （MT） engine, the present paper is intended to analyze the similarities and differences between Chinese and English in terms of SF, and to explore the general rules underlying the SF transfer of sentences from Chinese to English.