Modeling positive and negative feedback for improving document retrieval

作者：

Highlights：

• An improvement-based method is proposed to identify the feedback types.

• A novel negative feedback model which captures negative information is proposed.

• Integrating positive and negative feedback model is explored to improve retrieval.

• A content-based representativeness criterion is proposed and evaluated.

• Performance of the proposed models is evaluated on several TREC datasets.

摘要

•An improvement-based method is proposed to identify the feedback types.•A novel negative feedback model which captures negative information is proposed.•Integrating positive and negative feedback model is explored to improve retrieval.•A content-based representativeness criterion is proposed and evaluated.•Performance of the proposed models is evaluated on several TREC datasets.

论文关键词：Pseudo-relevance feedback,Negative feedback,Positive feedback,Language model

论文评审过程：Received 12 September 2018, Revised 8 November 2018, Accepted 26 November 2018, Available online 28 November 2018, Version of Record 30 November 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.11.035