Making an experimental corpus and its application to descriptive study of Japanese

本项目的目的是：(1)探索利用电子文本进行现代日语描述性研究的实用方法；(2)进行语料库实验。1.预备性研究(1)提出了考虑语篇体裁的平衡语料库规划的必要条件；(2)考察了现代日语语法研究论文中引用的例句，指出存在两组例句，即反思性例句和真实例句，这两组例句几乎是不相容的。(3)提出利用英日词典中的释义收集日语句型的建议。2.自20世纪90年代末S制作实验语料库以来，报纸和小说的电子文本在日语人文研究者中迅速流行起来。因此，我们收集了文本数据，这些数据在我们之间并不广泛共享。具体地说，我们收集了51部影视剧和其他文本的摘录。语料库的应用研究有效利用语料库的最好方法之一就是找出语法同义词之间的差异。我们考察了形容词后缀‘na’和‘no’，并给出了语义解释。

We have for the The aim of this project is (1)to explore practical method of making good use of electronic texts for descriptive study of modern Japanese and (2)to make a corpuse xperimentally.1.preparative study (1) We showed some necessary conditions for planning balanced-corpus considering text genre and in an essay (2) We investigated example sentences cited in the papers of grammatical study of modern Japanese and pointed out that there are two groups of showing example sentences which is in trospective group and real sample group and those two groups are almost incompatible. (3) We made an proposal to use definitions in English-Japanese dictionary for collecting Japanese sentence patterns.2.making experimental corpus Since late 1990's, electronic texts of newspapers and novels rapidly prevailed among humanities researchers of Japanese. Therefor we have collected text data which are not shared widely among us. To be specific we collected 51 scerarios of movie and TV drama and other texts. They made a total of approximately 4.8 mega bytes.3.applicative study of corpus One of the best way to use corpus effectively is to find out differences between grammatical synonymous expressions. We examined adnominal suffixes 'na' and 'no' and gave an semantical explanations.