论文部分内容阅读
本文提出了汉语信息处理中关于单字构词方式的基本问题 ,考察了目前对于这个问题的研究和应用情况。认为现有的统计性结论在未登录词处理中对于揭示单字构词的规律缺乏有效的作用。究其原因 ,一是这些结论体现的是词素组合成词之后的结构性质 ,而不是组合过程中的规律 ;二是这些调查统计遵循以句法为本的观点 ,而合成词的结构方式主要是意合。按照意合的构词观点 ,词素组合成词的过程要受多种语言要素和非语言因素的制约。目前还只能运用不完备的构词知识识别未登录词。文章最后给出了一组构词规则的工程化应用实例。
This paper presents the basic problems of word formation in Chinese information processing and examines the current research and application of this problem. It is considered that the existing statistical conclusion lacks an effective way to reveal the law of single word formation in the process of unregistered word processing. The reason is that these conclusions reflect the structural nature of morpheme combinations after the formation of words, rather than the law of the combination process; the second is that these survey statistics follow the syntax-based point of view, and the structure of the compound words mainly means Together According to the synopsis of the word viewpoints, morphemes into the word process by a variety of linguistic and nonverbal factors. At present, we can only use incomplete knowledge of word formation to recognize unregistered words. Finally, the article gives an example of engineering application of word formation rules.