Towards Improving English-Latvian Translation: A System Comparison and a New Rescoring Feature

This paper presents a comparative study of two alternative approaches to statistical machine translation (SMT) and their application to a task of English-to-Latvian translation. Furthermore, a novel feature intending to reflect the relatively free word order scheme of the Latvian language is proposed and successfully applied on the n-best list rescoring step. Moving beyond classical automatic scores of translation quality that are classically presented in MT research papers, we contribute presenting a manual error analysis of MT systems output that helps to shed light on advantages and disadvantages of the SMT systems under consideration.

• Type: Conference
• Year: 2010
• Where: LREC
• Authors: Maxim Khalilov, José A. R. Fonollosa, Inguna Skadina, Edgars Bralitis, Lauma Pretkalnina
• Keywords: Education, LREC 2010, MT Systems Output, Statistical Machine Translation, Word Order Scheme