LIG-CRIStAL System for the WMT17 Automatic Post-Editing Task
This work addresses automatic post-editing for machine translation, presenting incremental improvements in model performance for specific language pairs.
The paper tackled the WMT17 Automatic Post-Editing task by proposing two neural models: a monosource model with task-specific attention for low-resource scenarios and a chained architecture using source sentences for extra context, which slightly improved results with more training data, as tested on en-de and de-en datasets.
This paper presents the LIG-CRIStAL submission to the shared Automatic Post- Editing task of WMT 2017. We propose two neural post-editing models: a monosource model with a task-specific attention mechanism, which performs particularly well in a low-resource scenario; and a chained architecture which makes use of the source sentence to provide extra context. This latter architecture manages to slightly improve our results when more training data is available. We present and discuss our results on two datasets (en-de and de-en) that are made available for the task.