Verbosity normalized pseudo-relevance feedback in information retrieval
作者：
Highlights：
• We examine document length normalization on the pseudo-relevance feedback setting.
• We assume verbosity effect on term weights and scope effect on term selection.
• We generalize the existing two-stage normalization for the pseudo-relevance feedback.
• We apply the generalized two-stage normalization to the latent concept expansion.
• The resulting model improves the latent concept expansion on standard TREC datasets.
摘要
•We examine document length normalization on the pseudo-relevance feedback setting.•We assume verbosity effect on term weights and scope effect on term selection.•We generalize the existing two-stage normalization for the pseudo-relevance feedback.•We apply the generalized two-stage normalization to the latent concept expansion.•The resulting model improves the latent concept expansion on standard TREC datasets.
论文关键词：Pseudo-relevance feedback,Verbosity normalization,Scope normalization,Term frequency
论文评审过程：Received 4 April 2016, Revised 26 September 2017, Accepted 29 September 2017, Available online 23 November 2017, Version of Record 23 November 2017.
论文官网地址：https://doi.org/10.1016/j.ipm.2017.09.006
