Using Naïve Bayes Model and Natural Language Processing for Classifying Messages on Online Forum

In this paper, we would like to present how to classify Vietnamese messages on the online forum. We use the naive Bayes model for building a classifier. Besides, we also utilize natural language processing (NLP) tools for word segmentation, POS tagging, noun phrase chunking, extracting the nouns and noun phrases in message. These nouns and noun phrases are used for representing the message. With the representation model based on nouns and noun phrases, we improve the accuracy of classification with the support of semantic relations between words.