XMEANT: Better semantic MT evaluation without reference translations
XMEANT: Better semantic MT evaluation without reference translations
复制标题
DOI：
10.3115/v1/p14-2124
复制
发表时间：
2014-06
期刊：
影响因子：
--
通讯作者：
Chi-kiu (羅致翹) Lo;Meriem Beloucif;Markus Saers;Dekai Wu
中科院分区：
文献类型：
--
作者：
Chi-kiu (羅致翹) Lo;Meriem Beloucif;Markus Saers;Dekai Wu
我们引入了XMEANT--一种新的基于语义框架的机器翻译评价指标的跨语言版本--它与人类的充分性判断甚至比单一语言的手段更紧密地关联，并且消除了对昂贵的人类参考的需要。以前的工作确定了Means以最先进的准确性反映了翻译的充分性，针对Means优化机器翻译系统可以有力地提高翻译质量。然而，为了超越对数线性SMT模型中的权重调整，需要一个能够将语义框架标准深度整合到机器翻译训练管道中的跨语言目标函数。实验结果表明，跨语言XMEANT的性能优于单语MEANT，主要表现在：(1)用简单的翻译概率代替了单语语境向量模型；(2)加入了带括号的ITG约束。
We introduce XMEANT—a new cross-lingual version of the semantic frame based MT evaluation metric MEANT—which can correlate even more closely with human adequacy judgments than monolingual MEANT and eliminates the need for expensive human references. Previous work established that MEANT reflects translation adequacy with state-of-the-art accuracy, and optimizing MT systems against MEANT robustly improves translation quality. However, to go beyond tuning weights in the loglinear SMT model, a cross-lingual objective function that can deeply integrate semantic frame criteria into the MT training pipeline is needed. We show that cross-lingual XMEANT outperforms monolingual MEANT by (1) replacing the monolingual context vector model in MEANT with simple translation probabilities, and (2) incorporating bracketing ITG constraints.