Method for tagging part of speech and homograph, terminal device using the same

PURPOSE: A part of speech and homograph tagging method and a terminal device using the same are provided to use an HMM(Hidden Markov Model), thereby altogether reflecting contextual information to part of speech tagging and homograph tagging in natural language processing. CONSTITUTION: A word divider(110) divides a word of a text. A morpheme analysis string extractor(120) a morpheme analysis string from the divided word. A probability decision unit(160) decides the generating probability of the morpheme analysis string and transition probability among the morpheme analysis strings. A tagger(170) performs tagging for the morpheme analysis string of the word by using the generating and transition probability. The morpheme analysis string extractor extracts the morpheme analysis string corresponding to the word from a unigram database.