A new short-text categorization algorithm based on improved KSVM

A hybrid KSVM categorization algorithm is proposed in this paper, due to the fact that SVM algorithm classifies some tested temples in error nearby the optimal hyper-surface. In the classifying phase, the algorithm computes the distance from the tested sample to the optimal hyper-surface of SVM in the feature space, and chooses different algorithms for different distances. Then we apply this algorithm to short text categorization. The experimental results show that this algorithm, compared with traditional algorithms, greatly improved the classification accuracy of short text.