Paper / Neural Sign Language Translation

BLEU Score

BLEU (Bilingual Evaluation Understudy Score),is a metric used to evaluate a generated sentence to a reference sentence.

BLEU的范围是[0.0, 1.0],完美结果为1.0。BLEU的计算公式可以在[1]中找到。


ROUGE表示的是Recall-Oriented Understudy for Gisting Evaluation. 它是用在机器翻译和自动摘要中的评价指标。

ROUGE的范围是[0.0, 1.0], 完美结果为1.0。ROUGE的计算公式可以在[3]中找到。


[1] Statistical Machine Translation - Evaluation

[2] A Gentle Introduction to Calculation the BLEU Score for Text in Python

[3] ROUGE: A Package for Automatic Evaluation of Summaries