Sentence structure analysis based on semantic relations of cyrillic mongolian

There are many methods to do machine translation. Current experiments of machine translation systems usually based on statistics probability. However, it is not suitable to use the single method basing on statistics in terms of Mongolian language. Mongolian language is an agglutinative language. Thus, we assume that it is suitable to implement machine translation using method basing on inner essence in terms of Mongolian language. The analysis of the method based on language inner and outer sides are as follows: · Traditional grammar studied language on the part of form in surface structure level. It doesn´t rely on human brain characteristics. · And Universal grammar studied language on the part of meaning in deep structure level. It means language is not any external thing but internal thing of human brain. Consequently, we compared Mongolian grammar with Universal grammar method using the thesis and studied how to apply it in machine translation. To implement machine translation, we aimed at separating and choosing syntax parts which are relating to semantic relations from structural sentences. In order to do this, Mongolian subordinating relations and structural dependency are studied.