Abstract
As well-known, there are many syntactic similarities between Japanese and Korean language. These similarities allow us to build Japanese-Korean translation systems without depending too much on sophisticated syntactic analysis and semantic analysis. To further improve translation accuracy, we have been developing a Japanese-Korean translation system using these similarities for several years. However, there still remain some problems with regard to translation of inflected words, processing of multitranslatable words and so on. In this paper, we propose a new method for Japanese-Korean machine translation by using the relationships of two neighboring words. To solve the problems, we investigate the connection rules of auxiliary verb priority. And we design the translation table, which consists of entry tables and connection form tables. For unambiguous words, we can translate a Japanese word to the corresponding Korean word in terms of direct-matching method by consulting the only entry table. Otherwise, we have to evaluate the connection value computed from connection form tables and then we can select the most appropriate target word.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call