Complex Technology of Machine Translation Resources Extension for the Kazakh Language

Abstract

The paper is devoted to creating linguistic resources such as parallel corpora, dictionaries and transfer rules for machine translation for low resources languages. We describe the usage of Bitextor tool for mining parallel corpora from online texts, usage of dictionary enrichment methodology so that people without deep linguistic knowledge could improve word dictionaries, and we show how transfer rules for machine translation can be automatically learned from a parallel corpus. All describe methods were applied to Kazakh, Russian and English languages with a task of machine translation between these languages in mind.