Simple LSTM Weibo Text Classification

This is a simple text classification project through LSTM. The model used Keras framework in Python to train. This project will simply included two parts - Backend and Front-end.

The Backend folder included the trained model and the API server.

The Frontend folder included a webpage to call the api.

Dataset

• This dataset is downloaded from the Internet.

• The path of dataset will be:
  /Backend/data/

• The data included three categories that are design, health, and technology.

Train the model

• Running python keras_lstm.py to train the model

• The architecture will be Word Segmentation -> Embedding -> LSTM

• This model seem over fitting

• This trained model saved as:
  /Backend/keras_lstm.h5

Backend

• The Backend server will be using Flask

• The port of the API server will be 5002

• For running the server: python server.py

Front-End

• The Front-End will be written by simple Node.js throguh the Express framework

• The port of the webpage will be 8080

• For host the web server: node app.js