A default technique for the phonetic approach to retrieving Chinese characters
中文拼音文书处理的设定字输入技术
Sheng-ping Fang; Shyn-ming Chen; Kuang Mei 梅廣; Fu-wen Lin 林甫文; Jian-hua Huang

Abstract 摘要
This paper proposed a default technique to enhance the efficiency of Chinese text-processing devices using the phonetic alphabet. The principal idea of the default technique is to establish a context-sensitive table of default characters for a given syllable. For example, on receiving an input string li4, the computer would search for the most likely character pronounced as such and display it in the text area, so that the user could decide whether a correction is needed. The default character is determined by the transliteration of the target character and that of its preceding character. The default for a 1i4 preceded by /ju3/ could be 例, while that for li4 preceded by /mei3/ could be 丽. If the preceding context is not included in the default table, the most frequently used li4 (i.e., 力) would be displayed as the default. The rate of successful guessing, that is, the percentage of defaults matching targets, currently averaged 83%.

本文提出一种‘前后音设定字’输入技术以提高中文拼音文书处理的效率。此技术的基本观念是在电脑里存一套设定字字表。当使用者键入一个字音时（例如：li4），由电脑根据字表‘猜’他要的是那一个li4，而把设定的li4字显示出来，让使用者决定是否需要修改。至于li4字的设定，则参考文句里前一个音与li4的关系而决定。例如ju3li4的li4，设定字可能是‘例’，mei3li4的li4，设定字可能是‘丽’。如果文稿中li4的前一字字音不在设定字字表上，则单子使用率最高的li4（力）是设定字。目前设定字与目标字吻合的比率，亦即猜对率，平均为83%。

<< Back 返回

Readers 读者