In the paper, we use four popular LTR datasets: MQ2007, MSLR-WEB10K, Yahoo! and Istella-S. The experiment is divided into two parts: first, the Federated Learning phase needs to be conducted, followed ...