A New Probabilistic Model for Bayes Document Classification
Yashu Liu, Hanbing Yan · Proceedings of the 2nd International Conference on Computer Science and Electronics Engineering (ICCSEE 2013) · 2013
In this paper, we propose a new probabilistic model of naïve Bayes method which can be used in text classification.This method not only takes account of the frequency of feature words, but also considers those important words do not appear in the test document, which overcomes the shortcoming of the Multi-variate Bernoulli event Model(MBM) and Multinomial event Model(MM).Experiments show that the method proposed in this paper has better classification result than those traditional methods.