Constructing Financial Sentimental Factors in Chinese Market Using Techniques of Natural Language Processing

Introduction

Natural language processing, as one of the most promising fields of machine learning, has achieved great development recently and has been used in financial market. In this project, we are aiming to use an algotithm to analyze text data from influential financial websites to construct a sentimental factor which represents the daily sentiment of the market.
And papers here: English version and 中文版.

Experiment

Correlation Between Sentimental Factor and Chinese Markets

Time Series of Sentimental Factor and Chinese Markets

• As for SSE,

• As for SZSE,

Contribution

Contributors

• Junfeng Jiang
• Jiahao Li

Institutions

• AI&FintechLab of Likelihood Technology
• Sun Yat-sen University

Acknowledgement

We would like to say thanks to MingWen Liu from ShiningMidas Private Fund for his generous help throughout the research. We are also grateful to Xingyu Fu from Sun Yatsen University for his guidance and help. With their help, this research has been completed successfully.

Set up

Python Version

• 3.6

Modules needed

• six
• codec
• logging
• jieba
• gensim
• nltk
• selenium
• numpy
• pandas
• threading
• datetime

Contact

• [email protected]
• [email protected]
• [email protected]