Applying BERT to analyze investor sentiment in stock market
- PDF / 1,129,630 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 35 Downloads / 477 Views
(0123456789().,-volV)(0123456789(). ,- volV)
S.I. : HIGHER LEVEL ARTIFICIAL NEURAL NETWORK BASED INTELLIGENT SYSTEMS
Applying BERT to analyze investor sentiment in stock market Menggang Li1,3,4 • Wenrui Li2,3 • Fang Wang2,3 • Xiaojun Jia1,4 • Guangwei Rui2 Received: 21 August 2020 / Accepted: 29 September 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract This paper is an analysis of investor sentiment in the stock market based on the bidirectional encoder representations from transformers (BERT) model. First, we extracted the sentiment value from online information published by stock investor, using the Bert model. Second, these sentiment values were weighted by attention for computing the investor sentiment indicator. Finally, the relationship between investor sentiment and stock yield was analyzed through a two-step crosssectional regression validation model. The experiments found that investor sentiment in online reviews had a significant impact on stock yield. The experiments show that the Bert model used in this paper can achieve an accuracy of 97.35% for the analysis of investor sentiment, which is better than both LSTM and SVM methods. Keywords Bert model Investor sentiment Online reviews Cross-sectional regression
1 Introduction Many studies have shown that investor sentiment is an important factor in the price of stocks and other assets. From this perspective, changes in investor sentiment can predict market trends. Research shows that investor sentiment will be affected by news reports, and even the content, method, and frequency of news reports will affect investor sentiment. Therefore, investor sentiment can reflect investment willingness and expectations of market trends to a certain extent and has a certain ability to predict
& Wenrui Li [email protected] & Xiaojun Jia [email protected] 1
National Academy of Economic Security, Beijing Jiaotong University, Beijing, China
2
School of Economics and Management, Beijing Jiaotong University, Beijing, China
3
Beijing Laboratory of National Economic Security Earlywarning Engineering, Beijing Jiaotong University, Beijing, China
4
Beijing Center for Industrial Security and Development Research, Beijing Jiaotong University, Beijing, China
market yield, fluctuations, and transaction volume. It can be used as business operations, financial institution deposit, and loan decision-making and asset management, as well as new information channels for policy-making departments and regulatory agencies to conduct anticipated management. However, one of the most challenging problems in time series forecasting is generally considered to be stock market forecasting, due to its noisy and volatile nature. Another unresolved issue in modern socioeconomic and social organization is how to accurately predict stock movements. A lot of relevant research has emerged as an economic science research hotspot, coupled with the attraction of investor interest and high yields. Traditional stock market forecasting methods focus on time series a
Data Loading...
