Morphological interpretation method based on unregistered word type

摘要

The present invention relates to a morphological analysis method in Korean, and more particularly, to a morphological analysis method based on the type of an unregistered word. A morphological analysis method according to the present invention includes a preprocessing step of extracting words to be subjected to morphological analysis from a document, separating sentence codes, and processing numbers or special strings; A token generation step of generating tokens classified into Korean, numeric, English, and symbols for the result of the preprocessing; A candidate morpheme list generation step of generating all possible analysis candidates according to each token by referring to the frequency information and the part left and right access information, and performing a morphological analysis based on an unregistered word type and an estimation of a human name for a Hangul token; Selecting correct morpheme candidates with reference to frequency information and parts-of-speech left and right access information for a candidate morpheme list; And a post-processing process for processing language-dependent or specific language phenomena with respect to the result of the candidate morpheme selection process.;According to the morphological analysis method according to the present invention, it is possible to estimate the names of Korean people in advance, thereby reducing the occurrence frequency of unregistered words and improving the intelligibility of speech synthesis.