WORD SOUND DIVIDING SYSTEM FOR CHINESE WORD

摘要

PURPOSE:To automatically divide a word sound at a high speed and with high accuracy by applying a tone frequency method to divide the word tones of the Chinese sentences and therefore simplifying the dividing process algorithm. CONSTITUTION:A retrieving means for word tones, frequency classes and KANJI (Chinese characters) codes retrieves the data out of a dictionary and sends them to a clause point deciding means every time a syllable in inputted from a syllable input means (keyboard). The selection of an optimum division type is needed for the clause point deciding means only when =3 syllables are included. Then an alternative minimum frequency class sum deciding means of an optimum division generating means limits the optimum division type of the present time point to two candidates, i.e., a unisound or a bisound added at the end of a word tone string. Based on the difference between the frequency class sums of both tones, an optimum division type is designated. Thus a word tone string can be divided for each word sound with high accuracy.