Article Preview
TopIntroduction
The stock market is a crucial component of the market economy, playing a significant role in the stable operation and healthy development of the national economy (Shaban et al., 2024; Ali et al., 2023; Febrianti & Saadah, 2023; Amanda et al., 2023). Various trading data and derived technical indicators have been utilized for stock market prediction with the widespread popularity of stock trading. In particular, stock trend prediction has drawn extensive attention from scholars and investors, becoming a focal point and major research topic in the financial field (Gandhi et al., 2023; Juliani et al., 2023). Stock price movement is a nonlinear and nonstationary time series, not only following its own inherent laws but also being influenced by certain factors, such as culture, psychology, and policies, with obvious irrational characteristics (Kurani et al., 2023; Dash et al., 2023; Rejeb et al., 2023). Therefore, stock trend prediction is a challenging task (Garcia-Peñalvo et al., 2022). By actively learning and analyzing historical stock data, users can identify patterns and trends in price fluctuations, assess the risk and return characteristics of different stocks, validate the effectiveness of investment strategies, and understand market sentiment, thereby making better stock trend predictions (Rasmusen et al., 2022). There is a delicate balance between data sharing and privacy protection in stock trend prediction. On one hand, investors need to share their trading data and investment strategies to obtain more accurate prediction results. On the other hand, they also need to protect their privacy and avoid the disclosure of sensitive information (Xu et al., 2021). Currently, the main approach for stock trend prediction uses semantic web technologies to analyze and predict stock market data, aiding investors in making more informed investment decisions. The semantic web is a technology used for representing and sharing knowledge. It facilitates investors' understanding and processing of data by providing explicit semantic definitions to the data. This makes the semantic web highly suitable for stock trend prediction as it assists investors in better comprehending stock market data and identifying potential trends (Hu et al., 2022). In stock trend prediction, the chosen forecasting method that analyzes internal patterns of stocks and maximizes profits is a key factor (Brdesee et al., 2022). Current semantic web-based stock trend prediction methods can be primarily categorized into two types: traditional methods and machine learning algorithms (Rahayu & Ilham, 2023).
Traditional methods typically utilize techniques from economics and other fields to model the characteristics and technical indicators of stock trading for stock price prediction (Zhao et al., 2023; Sudipa et al., 2023). Typical methods include autoregressive moving average, autoregressive integrated moving average, and generalized autoregressive conditional heteroskedasticity models (Zhang, Li, et al., 2023; Yin et al., 2023; Sonkavde et al., 2023). However, the nonstationarity, nonlinearity, and other characteristics of stock prices result in significant limitations of traditional methods in stock trend prediction (Kurani et al., 2023; Dezhkam & Manzuri, 2023; Jiang et al., 2023; Amin et al., 2024).