Text representation for direction prediction of share market

作者：

Highlights：

• Study the relation between news articles and movement of close price of market.

• Combine news articles and prices using parallel CNN to predict the price movement.

• Experiment with different text representation schemes to represent news articles.

• News article text is more impactful in predicting than historical prices.

摘要

•Study the relation between news articles and movement of close price of market.•Combine news articles and prices using parallel CNN to predict the price movement.•Experiment with different text representation schemes to represent news articles.•News article text is more impactful in predicting than historical prices.

论文关键词：Financial time series,News articles,Price representation,Text representation,Deep learning

论文评审过程：Received 10 April 2022, Revised 15 July 2022, Accepted 5 August 2022, Available online 20 August 2022, Version of Record 30 August 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118472