Skip to content

Latest commit

 

History

History
14 lines (7 loc) · 710 Bytes

README.md

File metadata and controls

14 lines (7 loc) · 710 Bytes

Readme

This is a program based on machine learning which tries to classify the sentiment of product's comment in JD.

The accuracy of the classifier is about 85% in theory. It isn't good enough but the idea of the program may be useful.I trust that after optimization, this program will work better than now.

You can easily reuse that model only by writing DataSpider.py to get your data.

The detail has been shown in my CSDN blog.

基于机器学习的NLP情感分析(一)---- 数据采集与词向量构造方法

基于机器学习的NLP情感分析(二)—- 分类问题