Study on Chinese question classification based on SVM multi-category classification

Support vector machine (SVM) was initially used for binary classification. How to generalize the result of two-class classification to multi-class classification has been a problem which needs to be more investigated and studied. A general overview of existing representative methods for multi-category support vector machines was presented and their performances were compared in the paper. Then, the algorithms were applied in the Chinese question classification. Chinese question classification hierarchy and the feature selection of the question were also discussed in the paper. Then, The four algorithms of SVM multi-category classification were applied to Chinese question classification and some contrast experiments were done. The result of the experiments has shown that the binary-tree algorithm is more effective than the other algorithms in the Chinese question classification.