A text classification model based on training sample selection and feature weight adjustement

A new text classification model based on training samples selection and feature weight adjustment is presented. First it computes representativeness score of samples so as to distinguish noise samples from original training samples. Then a feature weight adjustment taking inter-class distribution and intra-class distribution into consideration is used to further improve the performance of text classification. The presented text classification model is applied on Chinese text dataset provided by Fudan Database Center. The experiments show that the proposed model can improve the performance of text classification to some extent with fewer training samples and fewer feature dimensions.