Collaborative community-specific microblog sentiment analysis via multi-task learning
作者：
Highlights：
• Using two sociological theories to extract social contexts of microblogs.
• Building a community-specific sentiment analysis model based on multi-task learning.
• Computing and incorporating community similarity as regularization over the model.
• Using social context regularization to share sentiment between connected microblogs.
• Optimizing the model based on a fast iterative shrinkage-thresholding algorithm.
摘要
•Using two sociological theories to extract social contexts of microblogs.•Building a community-specific sentiment analysis model based on multi-task learning.•Computing and incorporating community similarity as regularization over the model.•Using social context regularization to share sentiment between connected microblogs.•Optimizing the model based on a fast iterative shrinkage-thresholding algorithm.
论文关键词：Sentiment analysis,Microblogging,Multi-task learning,Social context
论文评审过程：Received 16 February 2020, Revised 3 July 2020, Accepted 13 November 2020, Available online 18 November 2020, Version of Record 10 February 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.114322
