Stenograph to Japanese Translation System

概要

論文の詳細を見る
A stenograph to Japanese (Kanji-Kana representation) translation system, which is one of the Japanese Information Processing systems, is introduced in this paper. Input data to this system are phonetic symbols expressing spoken Japanese and are not divided into Japanese words. The main feature of this system is that it translates the above-stated input string into normal Japanese sentences, and outputs them in combination of Chinese character, i.e. Kanji, and Kana alphabet. We have adopted a method of morphological analysis as translating algorithm and have made the dictionary by investigating spoken Japanese words. The mistranslating-ratio was 4% and the translating ratio of Chinese borrowed phonetic words into Kanji was 67%. The processing time was 155 msec/word.
一般社団法人情報処理学会の論文