Multi-factored and multi-classes text classification

We introduce several learning algorithm approaches for factored classification. In factored classification the class, label is factored into a vector of class features. A topic-based mixture model to factored classification of text documents in which each document is assumed to be generated by a mixture of class features is presented. Another three algorithms based on Naive Bayes model are compared with the topic-based mixture model. Experiments in factored text classification problems show that topic-based mixture model can outperform the other three approaches for categories with especially sparse training data.