Korean Part-of-Speech Tagging on a Hidden Markov Model