Qualitative Analysis On Morphological Features And Semantic Recognition Of Classical Chinese Function Words

As an isolated language,the appearance of Chinese words in different sentences has no obvious change.So we cannot figure out the grammatical framework of sentence just by word itself,which is entirely determined by function words and word order.There are sentences and phenomenon of classical Chinese in modern Chinese,because Ancient Chinese is the source of Modern Chinese.Study on the grammatical meaning and the pragmatic function of classical words,the establishment of knowledge base of classical words,and realization of semantic recognition of classical words,can make sense for effective theoretical basis for Chinese digital processing and intelligent search.According to the basic characteristics and word order of Chinese function words,firstly,we chose the word order as the main character,and the location and sentence patterns as supplementary characters to build the morphological features base.Secondly,sense text was transformed into symbolic calculable rules as the processing of extracting the morphological features.The code of function words,part of speech,grammatical component and phrase and sense rules were defined.According to the expert knowledge and the formal definition of rules,350 rules of meaning were constructed.Then we realized the rule labeling algorithm of function words in sentence,including a sentence pattern judgment algorithm for the four special sentence patterns,and word locating algorithm according to the relation between the function word form and the sentence position.Taking the matching results of morphological features extracted in the corpus and knowledge base,and using the artificial verification,the results show that the matching rules can better identify words semantic meaning in sentence.Meanwhile,we made some reasonable revisions in knowledge base for words which were failed to match the rules.