Exploiting syntactic and semantic information in coarse chinese question classification

Recent years have seen great process in studying English question classification. In our research, we learn Chinese question classification by exploiting the result of lexical, syntactic and semantic parsing on question sentences. Support vector machines are adopted to train a classifier on 6 coarse categories using single and combination of different parsing results as features. We find that even the surface information such as words and parts of speech could lead to a satisfying result, while augmenting the classifier with syntactic and semantic features could give even higher precision. However, the lack of words and incomplete syntactic structures among most questions cause combination of features even sparser than single features in the feature space, with much side effect brought to the performance of Chinese question classification.