Construcing Mongolian Attribute Information Database

This paper choose six thousands of simple sentences from the One Million Word Level Contemporary Mongolian Corpus, then tagged the fixed phrase, morphological features, semantic classification, sentence constituents and semantic roles, and developed construct the Mongolian attribute information database. Based on this database, we analysed the morphological features, the phrase classification and structural relation, the word number of phrase, the semantic classification of the head words in noun phrases and verb phrases, and the semantic roles.