Deep Domain Specialisation for single-model multi-domain learning to rank

TL;DR

This paper introduces Deep Domain Specialisation (DDS), a novel single-model approach for multi-domain learning in information retrieval that outperforms existing methods while reducing model complexity.

Contribution

The paper proposes DDS, a new architecture for consolidating multiple domains into one model, improving performance and efficiency over existing multi-domain models.

Findings

DDS outperforms Deep Domain Adaptation and baselines in experiments.

DDS requires fewer parameters per domain.

DDS demonstrates effectiveness in large-scale online testing.

Abstract

Information Retrieval (IR) practitioners often train separate ranking models for different domains (geographic regions, languages, stores, websites,...) as it is believed that exclusively training on in-domain data yields the best performance when sufficient data is available. Despite their performance gains, training multiple models comes at a higher cost to train, maintain and update compared to having only a single model responsible for all domains. Our work explores consolidated ranking models that serve multiple domains. Specifically, we propose a novel architecture of Deep Domain Specialisation (DDS) to consolidate multiple domains into a single model. We compare our proposal against Deep Domain Adaptation (DDA) and a set of baseline for multi-domain models. In our experiments, DDS performed the best overall while requiring fewer parameters per domain as other baselines. We show the efficacy of our method both with offline experimentation and on a large-scale online experiment on Amazon customer traffic.