|
Digital Assets Price Prediction Using Sentiment Analysis on Crowd Trading Idea |
|---|---|
| รหัสดีโอไอ | |
| Creator | Seksak Prabpala |
| Title | Digital Assets Price Prediction Using Sentiment Analysis on Crowd Trading Idea |
| Contributor | Kulthida Tuamsuk, Wirapong Chansanam |
| Publisher | Department of Information Science Faculty of Humanities and Social Sciences, Khon Kaen University |
| Publication Year | 2566 |
| Journal Title | Journal of Information Science |
| Journal Vol. | 41 |
| Journal No. | 3 |
| Page no. | 73-92 |
| Keyword | Digital assets, Crowd idea user, Sentiment analysis, Price prediction, Digital trading, Trading ideas |
| URL Website | https://www.tci-thaijo.org/index.php/jiskku/index |
| Website title | Journal of Information Science |
| ISSN | 2773-8841 |
| Abstract | "Purpose:The objective of this study is to conduct sentiment analysis of users' ideas on a social media platform, including comments about the prices of digital assets, to forecast potential price changes. Subsequently, the obtained results will be utilized to develop a decision support system for individual investors.Methodology: The study involves collecting ideas, trading techniques, graph analysis, and market opinions from online trader community, TradingView is the largest global community of cryptocurrency investors. Data was collected between September 1, 2022 and December 31, 2022, comprising a total of 8,725 text entries. The analysis focuses on categorizing the sentiments of these texts into three groups: positively related to a bullish market direction suggesting buying, negatively related to a bearish market direction suggesting selling, and neutral sentiments related to do not trading. Subsequently, a model is developed based on the majority user opinions to predict whether buying or selling is advisable during specific time periods.Findings: The forecasting accuracy using actual closing prices of digital assets compared to forecasting prices from various calculation methods, including 1-day overlap, standardized value, trend change analysis, percentage change, and correlation analysis, it was found that the actual prices and forecasting prices were in close agreement. The Pearson Correlation coefficient was as high as 0.89. Using the Granger-Causality Test, it was revealed that the sentiments of users on TradingView regarding digital asset prices tended to move in the same direction.Applications of this study: The analysis revealed a high correlation between actual and forecasted prices in the same direction. This result can be used to analyze turning points and forecast the prices of digital assets. By integrating the techniques from this research into an automated system that aggregates user opinions and analyzes real-time market sentiment, it can serve as a decision support tool for investment, including short-term price volatility predictions." |
