search

yao8839836 / text_gcn

Graph Convolutional Networks for Text Classification. AAAI 2019
1.36k stars 434 forks source link

最后的模型如何预测 #60

Open kangwenzhuang opened 5 years ago

kangwenzhuang commented 5 years ago

你好，读了您的paper，觉得很是厉害！但是有个地方不是很明白。如果有一个新的需要预测的测试集，是否还要重新加入节点重新跑代码计算，才能预测？关于预测我不是很懂。论文结论中的这段话我也不是很明白。However, a major limitation of this study is that the GCN model is inherently transductive, in which test document nodes (without labels) are included in GCN training. Thus Text GCN could not quickly generate embeddings and make prediction for unseen test documents. Possible solutions to the problem are introducing inductive (Hamilton, Ying, and Leskovec 2017) or fast GCN model (Chen, Ma, and Xiao 2018).谢谢！

yao8839836 commented 5 years ago

@kangwenzhuang

你好, Text GCN是需要将测试集加入图中重新训练才能预测，因为原始GCN需要将整个图的邻接矩阵A和节点特征矩阵X放入内存。测试节点和训练节点一起学习参数W0, W1, 然后得到embedding，最后用测试节点embedding预测标签。

原始GCN这种将有label的样本和要预测的样本放在一起学习并预测的方式称为transductive learning，这种方法不能对新样本进行“秒预测”。而GraphSAGE (Hamilton, Ying, and Leskovec 2017) 和FastGCN (Chen, Ma, and Xiao 2018) 通过采样的方式解决了以上问题，具体可以阅读这两篇论文。