A new hypersphere multi-class support vector machine applied in text classification

SVM is one of the most commonly used methods in the field of text classification. But, SVM is, in essence, a kind of binary classifier. When the traditional SVM is applied in text classification, many SVM must be trained, So the text classification accuracy is not ideal. In this paper, a new kind hypersphere support vector machine is applied in text classification, just require training a SVM. The SVM obtain a super ball center through training samples of each type text that in high-dimensional feature space, and then calculate the distance between the text sample to be tested and the center of each class, according to the minimum distance determine which class that the test text belongs to. The experimental results show that: with the measurement of Fl-measure the accuracy of the text classification has been greatly improved.