基于句子相似度的机器翻译评价方法及其有效性分析

为了帮助开发面向本地化的基于实例的机器翻译(EBMT)系统,提出利用句子相似度评价译文质量的方法．该方法能够较有效地区分不同翻译质量的译文,从而帮助开发者发现EBMT系统存在的问题．为了证明该评价方法的有效性,对比了6个机器翻译系统人工评价和自动评价的结果．通过相关分析和显著性检验,表明了此方法和人工方法存在高度的一致性．最后提出在自动评价结果的基础上利用线性回归预测人工评分方法．

Application and Analysis of String-Similarity-Based Machine Translation Evaluation

To help developing a localization-oriented example based machine translation (EBMT) system,an automatic machine translation evaluation method is implemented which adopts sentence similarity as criteria. Experiment shows that the evaluation method distinguishes well between "good" translations and "bad" ones. To verify the consistency between automatic and human evaluation methods,6 machine translation systems are scored using both methods and the evaluation results are compared. Correlation coefficient and significance tests are made to ensure the reliability of the results. Linear regression equations are calculated to map the automatic scoring results to human scorings,which can be utilized to predict human scoring of machine translation systems.