Character-level Convolution Neural Network for Text Classification in Keras

Train character-level convolutional network for text classification. Based on "Character-level Convolutional Networks for Text Classification" by Xiang Zhang, link here.

About this model

Attempt to build a machine can classify term through the text information found about that term. Using character-level as input minimizing the need of preprocessing, model could apply on different languages like Chinese.

Dependencies

• Python3
• Keras
• TensorFlow

Runtime model

• Runtime model pre-calculated all text. Result data build in Mongodb. Require PyMongo and a Mongodb server to run.