论文部分内容阅读
词类知识的粗糙和不完备是影响自动句法分析的一个重要因素。论文以一个完全按照语法功能标准建立起来的实词词类体系为依托,考察了基本名词短语的识别问题。研究表明,利用这个实词词类体系识别基本名词短语,正确率可以达到71.3%。
The rough and incomplete part-of-speech knowledge is an important factor that influences the automatic syntactic analysis. The dissertation relies on a real word class system which is established in full accordance with the grammatical function standard and examines the recognition of basic noun phrases. The research shows that using this real word phrase system to identify basic noun phrases, the correct rate can reach 71.3%.