Abstract
Predicting stock movements is a valuable research field that can help investors earn more profits. As with time-series data, the stock market is time-dependent and the value of historical information may decrease over time. Accurate prediction can be achieved by mining valuable information with words on social platforms and further integrating it with actual stock market conditions. However, many methods still cannot effectively dig deep into hidden information, integrate text and stock prices, and ignore the temporal dependence. Therefore, to solve the above problems, we propose a transformer-based attention network framework that uses historical text and stock prices to capture the temporal dependence of financial data. Among them, the transformer model and attention mechanism are used for feature extraction of financial data, which has fewer applications in the financial field, and effective analysis of key information to achieve an accurate prediction. A large number of experiments have proved the effectiveness of our proposed method. The actual simulation experiment verifies that our model has practical application value.
Highlights
In recent years, the use of social media information to predict the financial market has attracted the attention of more and more researchers, and some satisfactory experimental results have been achieved. is is due to the fact that social media information contains investor-related attitudes and subjective sentiments towards the financial market, resulting in many investment banks and hedge funds trying to dig out valuable information from media information to help better predict financial markets, which plays a key role in predicting the market
(x) CNN-BiLSTM-AM: this paper proposes a CNNBiLSTM-AM method to predict the stock closing price of the day, which is composed of convolutional neural networks (CNN), bidirectional long-short-term memory (BiLSTM), and an attention mechanism (AM) [18]
We have proposed a novel deep learning (DL) model called transformer encoder attention (TEA), which can use the historical stock prices from five calendar days in combination with social media tweet representations to predict stock movements by means of a transformer encoder and attention mechanisms
Summary
The use of social media information to predict the financial market has attracted the attention of more and more researchers, and some satisfactory experimental results have been achieved. is is due to the fact that social media information contains investor-related attitudes and subjective sentiments towards the financial market, resulting in many investment banks and hedge funds trying to dig out valuable information from media information to help better predict financial markets, which plays a key role in predicting the market. Complexity there are few studies using this method to predict the stock market; it is valuable to propose a research method that integrates social media information and stock prices. In order to effectively solve the above problems, we propose a novel transformer encoder attention (TEA) network architecture, which is a network framework for deep extraction of financial data features and further integration, including a feature extractor and a concatenation processor, where the historical data uses the information of five calendar days from the target trading day as the training data. (i) A network framework with a small-time window is proposed, which can effectively solve the problem of temporal dependence of financial data faced in the current stock movement prediction and realize the forecast with the help of historical data closer to the target trading day. We conclude the paper and identify future directions of research in the Conclusion and Future Work sections
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
