Japanese collocation extraction of information that takes into account the sentence structure

点击翻译按钮获取中文摘要

How to extract Japanese collocation information taking sentence structure into consideration was considered by this research. First, newspaper data and an actual newspaper checked what kind of differences there were. Next, newspaper data was corrected and two kinds of data, the data divided per one sentence and the data which has not been changed, were made. After the morphological analysis of each data, they were changed into the database and the collocation information on a high frequency vocabulary were extracted and the results were considered. Then, in order to divide BCCWJ per sentence, the sentence structure tags contained in BCCWJ were checked.