A Comparison of Neural Models for Word Ordering
This work addresses word ordering for natural language processing, presenting incremental improvements in model performance and efficiency.
The paper tackles the word-ordering task by proposing a new bag-to-sequence neural model with attention, which significantly outperforms existing models on a German WMT dataset and achieves better speed and quality on the English Penn Treebank.
We compare several language models for the word-ordering task and propose a new bag-to-sequence neural model based on attention-based sequence-to-sequence models. We evaluate the model on a large German WMT data set where it significantly outperforms existing models. We also describe a novel search strategy for LM-based word ordering and report results on the English Penn Treebank. Our best model setup outperforms prior work both in terms of speed and quality.