Reinforcement Learning to Rank in E-Commerce Search Engine: Formalization, Analysis, and Application

By Y. Hu et al

Published on June 10, 2018

Read the original document by opening this link in a new tab.

Table of Contents

1. Introduction
2. Background
3. Problem Formulation
4. Analysis of SSMDP

Summary

This paper focuses on the application of reinforcement learning to rank items in E-commerce search engines. It formalizes the multi-step ranking problem using a Search Session Markov Decision Process (SSMDP) and proposes a novel algorithm for learning an optimal ranking policy. The paper highlights the importance of maximizing accumulative rewards in a search session and presents experimental results demonstrating the superiority of the proposed algorithm over existing methods.