User and item-aware estimation of review helpfulness
作者：
Highlights：
• We study the impact on review helpfulness of author/item content deviations.
• We compare helpfulness prediction models based on linear and non-linear regression.
• Our models integrate author/item deviations with traditional helpfulness determinants.
• An experimental validation shows that our models outperform the selected baselines.
• We improve collaborative item recommendation using predicted review helpfulness.
摘要
•We study the impact on review helpfulness of author/item content deviations.•We compare helpfulness prediction models based on linear and non-linear regression.•Our models integrate author/item deviations with traditional helpfulness determinants.•An experimental validation shows that our models outperform the selected baselines.•We improve collaborative item recommendation using predicted review helpfulness.
论文关键词：Review helpfulness,Helpfulness determinants,Regression analysis,Helpfulness-aware personalized item recommendation
论文评审过程：Received 12 June 2020, Revised 21 September 2020, Accepted 4 November 2020, Available online 20 November 2020, Version of Record 20 November 2020.
论文官网地址：https://doi.org/10.1016/j.ipm.2020.102434
