Fig. 1.

The deep neural network architecture consists of a word embedding layer, followed by a convolutional layer with multiple filters, followed by a merge tensor, a fully connected dense layer and a single sigmoid output node