MultiWord Expression Aware Neural Machine Translation
Andrea Zaninello, Alexandra Birch
Abstract
Multiword Expressions (MWEs) are a frequently occurring phenomenon found in all natural languages that is of great importance to linguistic theory, natural language processing applications, and machine translation systems. Neural Machine Translation (NMT) architectures do not handle these expression well and previous studies have not explicitly addressed MWEs in this framework. In this work, we show that using external linguistic resources and data augmentation we can improve both translations of MWEs that occur in the source, and the generation of MWEs on the target, and improve performance by up to 5.09 BLEU points on MWE test sets. We also devise a MWE score to specifically assess the quality of MWE translation which agrees with human evaluation. We make available the MWEscore implementation – along with MWE-annotated training sets and corpus-based lists of MWEs – for reproduction and extension.