Knowledge empowered prominent aspect extraction from product reviews

作者：

Highlights：

• Prominent aspects are representatives of customer concerns for the given product.

• High-quality aspects cover most customer concerns with little semantic overlap.

• The linguistic knowledge and corpus statistics-based knowledge empower the extraction.

摘要

•Prominent aspects are representatives of customer concerns for the given product.•High-quality aspects cover most customer concerns with little semantic overlap.•The linguistic knowledge and corpus statistics-based knowledge empower the extraction.

论文关键词：Prominent aspect extraction,Unsupervised learning,Topic modeling,Word embedding

论文评审过程：Received 2 July 2018, Revised 15 October 2018, Accepted 19 November 2018, Available online 8 January 2019, Version of Record 8 January 2019.

论文官网地址：https://doi.org/10.1016/j.ipm.2018.11.006