Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2022-05-13
Abstract:
We explore how to crawl financial forum data such as stock bars and combine it with deep learning models for sentiment analysis. In this paper, we will use the BERT model to train against the financial corpus and conduct a comparative analysis of the Shenzhen stock index. By comparing the maximum information coefficients, it is found that the sentiment features obtained by applying the BERT model to the financial corpus can prove that the sentiment variables are correlated with the stock prices to a certain extent. Also this paper is an application of deep learning in a financial context. In further exploring the mechanism of investor sentiment on the stock market through the deep learning approach, it will be beneficial for national regulators and policy departments to formulate more reasonable policy guidelines on maintaining the stability of the stock market.
"
Peer Review Status:Awaiting Review
Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2022-05-10
Abstract: <p>"Based on the review data of Shenzhen stock index BBS from January 1, 2018 to December 31, 2019, this paper extracts the investor sentiment contained in it by using the deep learning BERT model, and studies the time-varying linkage relationship among investment sentiment, stock market liquidity and volatility by using TVP-VAR model. The experimental results show that investor sentiment has a stronger impact on the liquidity and volatility of the stock market, while the reverse impact is relatively small, but it changes more significantly with the state of the stock market. In addition, in all cases, the short-term response is more significant than that in the medium and long term, and the impact is asymmetric, and the impact in the market downturn is stronger.</p>
Peer Review Status:Awaiting Review