This repository accompanies the paper "A General Framework for Pairwise Unbiased Learning to Rank" published in ICTIR 2022. It contains code used for experiments described in the paper.
In order to recreate our experimental results, please follow the steps below.
- Install Docker.
- Change to the root directory of the project and run
./run.sh.
The script will start a docker container, perform simulations, and train and evaluate models.
The results will be produced in the form of summary tables saved in subdirectories under output.
Each subdirectory corresponds to a particular experimental setting
(defined by the click model and the maximum number of displayed positions). See the cited paper for more details.