Cognate-aware morphological segmentation for multilingual neural translation

Loading...
Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Conference article in proceedings
Date
2018-10-31
Major/Subject
Mcode
Degree programme
Language
en
Pages
8
390-397
Series
Third Conference on Machine Translation (WMT18); Brussels, Belgium
Abstract
This article describes the Aalto University entry to the WMT18 News Translation Shared Task. We participate in the multilingual subtrack with a system trained under the constrained condition to translate from English to both Finnish and Estonian. The system is based on the Transformer model. We focus on improving the consistency of morphological segmentation for words that are similar orthographically, semantically, and distributionally; such words include etymological cognates, loan words, and proper names. For this, we introduce Cognate Morfessor, a multilingual variant of the Morfessor method. We show that our approach improves the translation quality particularly for Estonian, which has less resources for training the translation model.
Description
| openaire: EC/H2020/780069/EU//MeMAD
Keywords
neural machine translation, morphology, cognate, multilingual
Other note
Citation
Grönroos , S-A , Virpioja , S & Kurimo , M 2018 , Cognate-aware morphological segmentation for multilingual neural translation . in Third Conference on Machine Translation (WMT18); Brussels, Belgium . Association for Computational Linguistics , pp. 390-397 , Conference on Machine Translation , Brussels , Belgium , 31/10/2018 . https://doi.org/10.18653/v1/W18-64037