Multi-Source Hard and Soft Information Fusion Approach for Accurate Cryptocurrency Price Movement Prediction (2409.18895v1)

Abstract: One of the most important challenges in the financial and cryptocurrency field is accurately predicting cryptocurrency price trends. Leveraging AI is beneficial in addressing this challenge. Cryptocurrency markets, marked by substantial growth and volatility, attract investors and scholars keen on deciphering and forecasting cryptocurrency price movements. The vast and diverse array of data available for such predictions increases the complexity of the task. In our study, we introduce a novel approach termed hard and soft information fusion (HSIF) to enhance the accuracy of cryptocurrency price movement forecasts. The hard information component of our approach encompasses historical price records alongside technical indicators. Complementing this, the soft data component extracts from X (formerly Twitter), encompassing news headlines and tweets about the cryptocurrency. To use this data, we use the Bidirectional Encoder Representations from Transformers (BERT)-based sentiment analysis method, financial BERT (FinBERT), which performs best. Finally, our model feeds on the information set including processed hard and soft data. We employ the bidirectional long short-term memory (BiLSTM) model because processing information in both forward and backward directions can capture long-term dependencies in sequential information. Our empirical findings emphasize the superiority of the HSIF approach over models dependent on single-source data by testing on Bitcoin-related data. By fusing hard and soft information on Bitcoin dataset, our model has about 96.8\% accuracy in predicting price movement. Incorporating information enables our model to grasp the influence of social sentiment on price fluctuations, thereby supplementing the technical analysis-based predictions derived from hard information.

• The paper introduces a multi-source fusion approach that integrates historical market data with social sentiment analysis.
• It utilizes a BiLSTM framework combined with FinBERT to process technical indicators and qualitative data effectively.
• Empirical tests on Bitcoin data show a prediction accuracy of approximately 96.8%, outperforming single-source models.

The paper "Multi-Source Hard and Soft Information Fusion Approach for Accurate Cryptocurrency Price Movement Prediction" addresses the challenge of predicting cryptocurrency price trends by leveraging a novel approach that combines hard and soft information sources. This fusion aims to enhance the accuracy of such forecasts, particularly in volatile and data-rich cryptocurrency markets.

Hard and Soft Information Fusion (HSIF) Approach

Hard Information:

• Historical Price Records: Includes past prices of cryptocurrencies.
• Technical Indicators: Metrics such as moving averages and relative strength index (RSI) which are used to interpret financial data.

Soft Information:

• Social Media and News Data: Extracted from X (formerly Twitter), including tweets and news headlines about cryptocurrencies.
• Sentiment Analysis: Employs a BERT-based model, specifically FinBERT, due to its superior performance in financial contexts. This model analyzes the sentiment of the soft data, categorizing it to understand the market sentiment better.

Model Framework

The proposed model integrates both types of information into a framework where the processed data is used to predict cryptocurrency price movements. The data is fed into a bidirectional long short-term memory (BiLSTM) model. The BiLSTM is chosen because of its ability to process data in both forward and backward directions, capturing long-term dependencies in sequential information.

Empirical Findings

The paper primarily tests the model on Bitcoin-related data and achieves an impressive accuracy of approximately 96.8% in predicting price movements. The findings emphasize that the HSIF approach significantly outperforms models that rely on single-source data. This high accuracy rate showcases the model's ability to incorporate the influence of social sentiment, gleaned from the soft data, into the technical analysis provided by the hard data.

Conclusion

By merging hard and soft information, the HSIF approach provides a comprehensive framework that accounts for both the quantifiable data from market histories and the qualitative data from social sentiments. This fusion allows the model to grasp the complex dynamics of cryptocurrency price movements more effectively than traditional single-source models. The integration of FinBERT for sentiment analysis and the use of BiLSTM for processing the fused dataset are crucial facets contributing to the enhanced predictive performance showcased in this paper.