Pairwise learning to rank by neural networks revisited: reconstruction, theoretical analysis and practical performance


Loading...

Date

2025-04

Publication Type

Journal Article

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

We reevaluate the pairwise learning to rank approach based on neural nets, called RankNet, and present a theoretical analysis of its architecture. We show mathematically that the model can, under certain conditions, learn reflexive, antisymmetric, and transitive relations, enabling simplified training and improved performance. Experimental results on the LETOR MSLR-WEB10K, MQ2007 and MQ2008 datasets show that the model outperforms numerous state-of-the-art methods (including a listwise approach), while being inherently simpler in structure and using a pairwise approach only.

Publication status

published

Editor

Book title

Volume

114 (4)

Pages / Article No.

112

Publisher

Springer

Event

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Information retrieval; Machine learning; Learning to rank

Organisational unit

Notes

Funding

Related publications and datasets