An NLP-PCA Based Trading Strategy On Chinese Stock Market

DOI

10.2991/hsmet-19.2019.16How to use a DOI?

Keywords

Chinese stock analysis, Financial sentiment analysis, NLP, PCA.

Abstract

The stock market is a barometer of a country's economy. However, the stock market is significantly affected by policies, news, and public opinion, and it is prone to volatility. Compared with the already mature financial securities market in foreign countries, China's stock market is still in the exploratory stage. There are more individual stock speculators in short-term trading. They will search for news through various channels to make decisions. Behavioral finance has created a theoretical basis for the mining of stock reviews. The rise of technologies such as text mining, machine learning, and time series models has made stock review mining possible. In this paper, we extract 28 kinds of financial sentiment features from thousands of Chinese news and social media outlets by NLP. Compared with popular methods where they only use positive or negative sentiment to predict stock price, we find out five more specific information categories from the news, which is SSE rising or dropping expectation, macro-public finance, bond sentiment, bond price forecast, the buzz on bond, rate and stock index. Finally, we predict stock price using these primary factors, providing a specific predictive ability to the stock market trend.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Zhao Liu
AU  - Huiying Zhu
AU  - Tham Yew Chong
PY  - 2019/07
DA  - 2019/07
TI  - An NLP-PCA Based Trading Strategy On Chinese Stock Market
BT  - Proceedings of the 4th International Conference on Humanities Science, Management and Education Technology (HSMET 2019)
PB  - Atlantis Press
SP  - 80
EP  - 89
SN  - 2352-5398
UR  - https://doi.org/10.2991/hsmet-19.2019.16
DO  - 10.2991/hsmet-19.2019.16
ID  - Liu2019/07
ER  -