Blending Search and Discovery: Tag-Based Query Refinement with Contextual Reinforcement Learning

Abstract

We tackle tag-based query refinement as a mobile-friendly alternative to standard facet search. We approach the inference challenge with reinforcement learning, and propose a deep contextual bandit that can be efficiently scaled in a multi-tenant SaaS scenario.

arXiv commentaccepted at EComNLP 2020

References in corpus (2)

• A Contextual Bandit Bake-off
• Taming Non-stationary Bandits: A Bayesian Approach

Cited by in corpus (1)

• Query2Prod2Vec Grounded Word Embeddings for eCommerce