Personalized sentiment classification of customer reviews via an interactive attributes attention model
作者：
Highlights：
• A bilinear term was used to capture interactive features between attributes and text.
• A multiloss objective function was applied on all four different representations.
• Comparative experiments demonstrate the proposed model outperforms prior methods.
摘要
•A bilinear term was used to capture interactive features between attributes and text.•A multiloss objective function was applied on all four different representations.•Comparative experiments demonstrate the proposed model outperforms prior methods.
论文关键词：Sentiment analysis,Text classification,Sentence representation,Bilinear attention
论文评审过程：Received 21 November 2020, Revised 28 February 2021, Accepted 11 May 2021, Available online 12 May 2021, Version of Record 18 May 2021.
论文官网地址：https://doi.org/10.1016/j.knosys.2021.107135
