A Deep Look into neural ranking models for information retrieval

作者：

Highlights：

• We introduce some typical IR tasks addressed by neural ranking models and discuss major characteristics and challenges.

• We introduce a unified formulation over neural ranking models, and review existing models based on this formulation.

• We survey published empirical results on the ad-hoc retrieval and QA tasks to conduct a comprehensive comparison.

• We discuss several trending topics that are important or might be promising in the future.

摘要

•We introduce some typical IR tasks addressed by neural ranking models and discuss major characteristics and challenges.•We introduce a unified formulation over neural ranking models, and review existing models based on this formulation.•We survey published empirical results on the ad-hoc retrieval and QA tasks to conduct a comprehensive comparison.•We discuss several trending topics that are important or might be promising in the future.

论文关键词：Neural ranking model,Information retrieval,Survey

论文评审过程：Received 9 March 2019, Revised 8 June 2019, Accepted 20 June 2019, Available online 9 July 2019, Version of Record 20 October 2020.

论文官网地址：https://doi.org/10.1016/j.ipm.2019.102067