Analyzing the structural features of Mongolian sentence constituents deepens the knowledge of the structural features of the sentence constituents and provides the formal knowledge to the computer recognition of the sentence constituents. At the first, we choose 5107 sentences from the real text corpus including about 2000 thousand words.
Then, we tag the morphological and syntactic features on them. At the end, by statistically analyzing the results of the tagging, we found that the NP can be the object, the subject, the adverbial, and the attribute in the sentences with the verb-predicates, in which the objects appear most often while the attributes appear with the least number. In order to describe the features of NP subject of sentence, we also statistically analyze their POS structures and the corresponding morphological information.