This paper introduces a novel Graph Convolutional Neural Network (GCN) model with dense connections and an attention mechanism for text classification. The proposed model, named DC-GCN, addresses the limitations of existing GCN models by increasing the depth of the GCN network, which allows it to capture long-distance node information and reflect various scale features of the text. The dense connections in the DC-GCN multiplex small-scale features from shallow layers, generating different scale features. An attention mechanism is also incorporated to combine these features and determine their relative importance. Experimental results on four benchmark datasets (MR, R8, R52, and Ohsumed) demonstrate that the DC-GCN model outperforms conventional deep learning text classification models and other GCN algorithms, particularly for short and medium-length texts. The model's effectiveness is attributed to its ability to extract more effective semantic features and its efficient handling of text features at different scales.This paper introduces a novel Graph Convolutional Neural Network (GCN) model with dense connections and an attention mechanism for text classification. The proposed model, named DC-GCN, addresses the limitations of existing GCN models by increasing the depth of the GCN network, which allows it to capture long-distance node information and reflect various scale features of the text. The dense connections in the DC-GCN multiplex small-scale features from shallow layers, generating different scale features. An attention mechanism is also incorporated to combine these features and determine their relative importance. Experimental results on four benchmark datasets (MR, R8, R52, and Ohsumed) demonstrate that the DC-GCN model outperforms conventional deep learning text classification models and other GCN algorithms, particularly for short and medium-length texts. The model's effectiveness is attributed to its ability to extract more effective semantic features and its efficient handling of text features at different scales.