Study of a Multilingual Translation Process Using Statement Restriction Rules

التفاصيل البيبلوغرافية
العنوان: Study of a Multilingual Translation Process Using Statement Restriction Rules
المؤلفون: Yuu Nakajima, Reiko Hishiyama, Yoshiyuki Kitajima, Kana Tajima
المصدر: ACIT/CSII/BCD
بيانات النشر: IEEE, 2017.
سنة النشر: 2017
مصطلحات موضوعية: Statement (computer science), Machine translation, Process (engineering), business.industry, Computer science, media_common.quotation_subject, computer.software_genre, Terminology, Quality (business), Artificial intelligence, business, computer, Natural language processing, Range (computer programming), media_common, Meaning (linguistics)
الوصف: Recently, the number of foreigners living in or visiting Japan has increased. Foreigners in Japan use a wide range of languages, and handling all of these languages is difficult. Although machine translation can be used to provide economical translations, cases exist in which this technology does not generate sentences that properly communicate the meaning of the original text depending on the source language. This research proposes a translation process that applies statement restriction rules to create sentences that can be easily machine translated in order to improve the quality of multilingual translation. It is based on the hypothesis that statement restriction rules meant to improve the translation quality of a given language pair (such as Japanese to English) will also improve machine translation quality for other language pairs. In this research, we create Japanese statement restriction rules based on Japanese-English translations, and demonstrate that the proposed translation process can be effective in translating Japanese sentences created with these statement restriction rules into English, Chinese, and Korean.
URL الوصول: https://explore.openaire.eu/search/publication?articleId=doi_________::c9bb6a1b33cedfe5f420daaa72c93b5a
https://doi.org/10.1109/acit-csii-bcd.2017.22
رقم الأكسشن: edsair.doi...........c9bb6a1b33cedfe5f420daaa72c93b5a
قاعدة البيانات: OpenAIRE