Translating from Original to Simplified Sentences using Moses: When does it Actually Work?

In recent years, several studies have approached the Text Simplification (TS) task as a machine translation (MT) problem. They report promising results in learning how to translate from ‘original’ to ‘simplified’ language using the standard phrasebased translation model. However, our results indicate that this approach works well only when the training dataset consists mostly of those sentence pairs in which the simplified sentence is already very similar to its original. Our findings suggest that the standard phrase-based approach might not be appropriate to learn strong simplifications which are needed for certain target populations.