Dynamic System for Multi-Factor Cryptocurrency Forecasting: Integrating Emotion Weighting and Herding Effects on Binance Top 10

¹ Shanghai Mingsui Creative School, 200120 Shanghai, China
² Albert College, Belleville City, Canada
³ School of Ethnology and Sociology, Minzu University of China, 100000 Beijing, China

^* Corresponding author’s email: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Cryptocurrency markets are notorious for their extreme volatility and social sentiment susceptibility, making them challenging for accurate price forecasting. To tackle these problems, this paper proposes Dynamic Sentiment Engine– a predictive model for the top ten cryptocurrencies on Binance including Bitcoin and Ethereum. The system proposes a multi-dimensional solution which incorporates real-time multi-source social media sentiment analysis results and extends herd behavior tracking to financial markets. The innovation lies in the adaptive weighting scheme which adjusts the relative influence between collective investor sentiment and emerging herd dynamics in a market-regime calibrated way. By continuously adapting the relative influence between sentiment and herd effects in stable or turbulent market regimes, the prediction robustness is greatly improved over traditional univariate or static models. The system provides traders with a novel decision-support tool to operate in the highly volatile digital asset market, and contributes to behavioral finance by providing a measurable model to explore sentiment-herd interdependencies.