Uyghur Short Text Classification Using Morphological Information

In this paper, we propose a novel method for improving the classification performance of short text strings using conditional random fields (CRFs) that combine morphological information. Experimental results on three datasets (Uyghur, Chinese, and English) demonstrate that our method can yield higher classification accuracy than Support Vector Machine (SVM) classifier and Maximum Entropy Model (MEM) classifier. Moreover, we show that our method can greatly decrease error rates, particularly if the number of training texts or the size of the strings in the train set is small.