Leveraging Contextual Sentences for Text Classification by Using a Neural Attention Model

<div>The two basic models in our experiments. One is based on CNN and the other is based on BLSTM. (a) The basic model based on CNN. The width of convolution filters is 3, and the method of pooling is max pooling. (b) The basic model based on LSTM. The output of LSTM network is a matrix concatenating the hidden vectors from forward LSTM and backward LSTM.</div>

Computational Intelligence and Neuroscience

fig1

Figure 1

Figure 1: Leveraging Contextual Sentences for Text Classification by Using a Neural Attention Model 