In this study, we present the NER systems based on word and syllable levels as well as compare the performances of these two systems. The results show that their performances
are not much different from each other. However, as the features used in this experiment are only the word lists, the dictionaries, unigram and bigram, the systems consequently do not perform well enough. For further studies, we plan to apply more features such as context clues in our systems and increase the training data in order to improve the systems’ performances.