MACHINE TRANSLATION METHOD FOR SEPARATING COMPOUND NOUN USING MARKOV MODEL

摘要

PURPOSE: A machine translation method is provided to separate a compound noun by statistically analyzing the compound noun and designing and learning a Markov model thereof. CONSTITUTION: The machine translation method includes a morpheme analyzing step, a tagging step, an idiom processing step, a vocabulary pre-processing step, a conversion rule application step, a meaning vagueness solving step, and an object language generating step. A compound noun is separated to process the compound noun registered in the vocabulary pre-processing step. The compound noun separating step includes a mechanical learning step and a separating step. In the separating step, the compound noun is separated from using a separating algorithm by using a mechanical learning data of the mechanical learning step.