摘要： 提出了一种改进的汉语句子相似度衡量方法,用于基于实例的汉英机器翻译。该方法同时考虑了相同词串的数目及长度和对应的权值信息,克服了传统方法的显著不足,在理论上更有合理性。在小数据集上的实验也表明该方法是可行的。

关键词: 基于实例机器翻译, 汉英机器翻译, 句子相似度衡量, 自然语言处理

Abstract: This paper proposes an improved Chinese sentence similarity measure for example-based Chinese to English translation. The length of identical word sequence and the weights of words in the sequence are considered in this method. The preliminary experimental results show that the measure work is well when it is tested on a small dataset.

中图分类号: 

• TP391.1