Learning to rank products based on online product reviews using a hierarchical deep neural network
作者：
Highlights：
• A hierarchical approach is proposed to product ranking (specifically, word level and review level) to provide more supervision.
• A hierarchical attention network is extended to operate in the ranking domain with learning-to-rank strategies.
• Results reveal that the hierarchical architecture of the ranking model improves the performance of sales rank predictions in Amazon Product Data.
• The attention weights in the ranking model can be used to analyze which content has a significant impact on the ranking of products.
摘要
•A hierarchical approach is proposed to product ranking (specifically, word level and review level) to provide more supervision.•A hierarchical attention network is extended to operate in the ranking domain with learning-to-rank strategies.•Results reveal that the hierarchical architecture of the ranking model improves the performance of sales rank predictions in Amazon Product Data.•The attention weights in the ranking model can be used to analyze which content has a significant impact on the ranking of products.
论文关键词：Product ranking,Online product reviews,Hierarchical deep neural network
论文评审过程：Received 28 December 2018, Revised 17 May 2019, Accepted 30 June 2019, Available online 2 July 2019, Version of Record 13 July 2019.
论文官网地址：https://doi.org/10.1016/j.elerap.2019.100874
