摘要
本文提出并讨论了拼音汉字转换输入法中的常用词搭配结构的识别方法。该方法在词组的水平上,在常用搭配结构约束下,利用少量的词属性特征,可有效的解决许多单字词的识别问题,这其中包括一些单纯句法难以识别的同音单字词。
The structure recognition of word collocation in the Pinyin-Chinese transformation for Chinese input is proposed, and five different commonly used word collocation structures have been discussed inthis paper. Confined in the relationship of word collocation and analysed at the level of compound word structure, this method could effectively overcome many difficulties in homophone word recognition by using only the information of some simple word attributes.
出处
《中文信息学报》
CSCD
1992年第1期25-32,共8页
Journal of Chinese Information Processing
