Short text classification based on convolutional upsampling feature enhancement

Short text can lead to sparse feature representation and classification inaccuracies due to noise and other issues. To address this, we propose a short text classification model that uses convolutional upsampling feature enhancement. Our approach involves using a multi-scale convolutional neural network to extract deep features of different dimensions. Secondly, we enhance the deep features by upsampling convolution and obtain more discriminative features for text classification by downsampling. Finally, we use an end-to-end approach to output the text categories. Experimental validation on a public dataset shows that our proposed feature enhancement approach significantly improves the model's performance in short text categorization.