MACHINE TRANSLATION SYSTEM WITH FUNCTION LEARNING DESIGNATION OF PARTS OF SPEECH BY USERS

摘要

PURPOSE:To improve the correct answer rate of the estimation of a part of speech when it is difficut to judge whether either one is correct in a grammar rule by storing/using the parts of speech information designated by users in a machine translation system provided with a function learning the designation of parts of speech by users. CONSTITUTION:An input part 1 receives an original sentence. An analysis part 2 analyzes the original sentence by obtaining vocabulary information from a dictionary 3. A generation part 4 performs the generation of a translation by referring to the dictionary 3. An output part 5 displays the parts of speech information of each word of the translation and the original sentence to users. A part of speech designation part 6 receives the parts of speech information designated by users and passes it to the input part 1 and a part of speech learning file 7. The parts of speech information passed to the part of speech learning file 7 is recorded so that it may be used for the coming parts of speech estimation processing.