'From-Bottom-to-top' to Analyze Sentence Constituent of Traditional Mongolian Basing on The Rule

摘要

Syntactic Parsing has played an important role in Natural Language Processing (NLP).The character of Traditional Mongolian is that "the predicate is generally at the end of the sentence, the other constituent can change location but the meaning of the sentence is not change". According to this character, the paper propose a "from-bottom-to-top" method to analyze the sentence constituent. Part-of-Speech (POS) tagging is the first step of analyzing the sentence constituent. Marking POS of words and phrases based on the dictionary library and the rule base. After the preprocessing, the sentence is divided into several modules by keywords, "case", phrase, and so on. Every module use the "from-bottom-to-top" method to analyze and label the sentence constituent.