COVID-19 personal health mention detection from tweets using dual convolutional neural network

作者：

Highlights：

• We built and released the first COVID-19 tweet dataset for PHM identification.

• We proposed a dual CNN network to classify each tweet into four PHM categories.

• The P-Net and A-Net in the dual CNN well alleviate the class imbalance issue.

• State-of-the-art performance was achieved.

摘要

•We built and released the first COVID-19 tweet dataset for PHM identification.•We proposed a dual CNN network to classify each tweet into four PHM categories.•The P-Net and A-Net in the dual CNN well alleviate the class imbalance issue.•State-of-the-art performance was achieved.

论文关键词：CNN,Deep learning,Health monitoring,Social media,Text mining

论文评审过程：Received 3 June 2021, Revised 13 January 2022, Accepted 29 March 2022, Available online 2 April 2022, Version of Record 5 April 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117139