PostCom2DR: Utilizing information from post and comments to detect rumors
作者：
Highlights：
• We propose a rumor detection model only based on the post and comments.
• Reply structure of post and comments is modeled to detect rumors.
• Mutual selection between post and comments is considered in our work.
• Topic drift within the comments is utilized to rumor detection.
• The detection ability of our model is remarkable.
摘要
•We propose a rumor detection model only based on the post and comments.•Reply structure of post and comments is modeled to detect rumors.•Mutual selection between post and comments is considered in our work.•Topic drift within the comments is utilized to rumor detection.•The detection ability of our model is remarkable.
论文关键词：Rumor detection,Reply structure of post and comments,Graph convolution network,Post-comment co-attention,CNN-based local representation
论文评审过程：Received 15 March 2021, Revised 13 June 2021, Accepted 10 October 2021, Available online 27 October 2021, Version of Record 3 November 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.116071
