Introduction to the

The article classification model trained by Text convolutional Neural network (TextCNN) is used to detect whether the article is pornographic.

Author:yudake
date:2018/2/22
Copy the code

See this blog post for a detailed breakdown of the project

Development environment:

  • Python3. X;
  • Tensorflow1.2;
  • Jieba;
  • Other related class libraries will be mentioned in the program.

Objective: To achieve pornography detection.

Model effect: The final accuracy rate is above 98%.

model

  • Input: sentence matrix of the article after processing;
  • Model: Text convolutional neural network;
  • Output: classification results.

Accuracy curve

The average accuracy was 98%. See the code and blog for details.