Machine Translation Evaluation using Recurrent Neural Networks

Machine Translation Evaluation using Recurrent Neural Networks
复制标题

DOI:
10.18653/v1/w15-3047
复制
发表时间:
2015-09
期刊:
--
影响因子:
--
通讯作者:
Rohit Gupta;Constantin Orasan;Josef van Genabith
Rohit Gupta;Constantin Orasan;Josef van Genabith
中科院分区:
其他
文献类型:
--
作者:
Rohit Gupta;Constantin Orasan;Josef van Genabith

文献摘要

被引文献

相似文献

本文介绍了我们的度量(UoWLSTM)提交的WMT-15度量任务。许多最先进的机器翻译(MT)评估指标都很复杂,涉及大量的外部资源(例如用于释义),并且需要进行调整以实现最佳结果。我们使用基于稠密向量空间和长短期记忆(LSTM)网络的度量,它们是递归神经网络(RNN)的类型。对于WMT 15,根据斯皮尔曼和Pearson(Pre-TrueSkill),我们的新指标是整体性能最好的指标,根据Pearson(TrueSkill)系统级相关性,我们的新指标是第二好的指标。
This paper presents our metric (UoWLSTM) submitted in the WMT-15 metrics task. Many state-of-the-art Machine Translation (MT) evaluation metrics are complex, involve extensive external resources (e.g. for paraphrasing) and require tuning to achieve the best results. We use a metric based on dense vector spaces and Long Short Term Memory (LSTM) networks, which are types of Recurrent Neural Networks (RNNs). For WMT15 our new metric is the best performing metric overall according to Spearman and Pearson (Pre-TrueSkill) and second best according to Pearson (TrueSkill) system level correlation.