基于BERT模型的投资者情绪指数建模及与价格关系分析

Modeling of Investors’Sentiment Index Based on BERT Model and Analysis of Its Relationship with Price

基于BERT模型,应用21家期货公司行情预测分析文本数据,构建了期货市场投资者情绪指数;在此基础上,运用格兰杰因果检验分析了期货市场价格与市场情绪指数的相互影响作用。研究结果表明,BERT模型相较基于经典分类算法模型在各评价指标上均有约10%的提升。同时,投资者情绪指数与期货收盘价之间存在相互影响,期货收盘价对投资者情绪的影响程度更大,影响持续时间更短。

Based on the BERT model,the text data of 21 futures companies’ market prediction and analysis was used to construct an investors’ sentiment index for the futures market.On this basis,Granger causality test was used to analyze the interaction between the futures market price and the market sentiment index.The research results show that the BERT model has improved about 10%in various evaluation indicators compared with the model based on the classic classification algorithm.At the same time,there is a mutual influence between the investors’ sentiment index and the futures closing price.The futures closing price has a greater degree of influence on investors’ sentiment and has a shorter duration.