The paper "Graph Convolutional Networks for Text Classification" by Liang Yao, Chengsheng Mao, and Yuan Luo proposes a novel approach to text classification using graph convolutional networks (GCNs). The authors construct a large heterogeneous text graph that captures both global word co-occurrence and document-word relations. They then apply a two-layer GCN to learn node embeddings for words and documents, which are used as input to a softmax classifier for text classification. The method is evaluated on multiple benchmark datasets, demonstrating superior performance compared to state-of-the-art methods without using pre-trained word embeddings or external knowledge. The experimental results show that Text GCN can effectively learn predictive word and document embeddings and perform well even with limited labeled data. The paper also discusses the limitations of the approach, such as its transductive nature, and suggests future directions for improving the model's performance and scalability.The paper "Graph Convolutional Networks for Text Classification" by Liang Yao, Chengsheng Mao, and Yuan Luo proposes a novel approach to text classification using graph convolutional networks (GCNs). The authors construct a large heterogeneous text graph that captures both global word co-occurrence and document-word relations. They then apply a two-layer GCN to learn node embeddings for words and documents, which are used as input to a softmax classifier for text classification. The method is evaluated on multiple benchmark datasets, demonstrating superior performance compared to state-of-the-art methods without using pre-trained word embeddings or external knowledge. The experimental results show that Text GCN can effectively learn predictive word and document embeddings and perform well even with limited labeled data. The paper also discusses the limitations of the approach, such as its transductive nature, and suggests future directions for improving the model's performance and scalability.