Automatic Expert Selection for Multi-Scenario and Multi-Task Search

TL;DR

This paper introduces AESM^{2}, an automatic expert selection framework that unifies multi-scenario and multi-task learning with hierarchical structure and dynamic expert selection, improving performance in search systems.

Contribution

AESM^{2} is a novel framework that automatically learns the structure and expert selection for multi-scenario and multi-task search, overcoming limitations of static architectures like MMoE.

Findings

AESM^{2} outperforms strong baselines on large-scale datasets.

Online A/B testing shows substantial performance improvements.

AESM^{2} has been deployed online for major traffic serving.

Abstract

Multi-scenario learning (MSL) enables a service provider to cater for users' fine-grained demands by separating services for different user sectors, e.g., by user's geographical region. Under each scenario there is a need to optimize multiple task-specific targets e.g., click through rate and conversion rate, known as multi-task learning (MTL). Recent solutions for MSL and MTL are mostly based on the multi-gate mixture-of-experts (MMoE) architecture. MMoE structure is typically static and its design requires domain-specific knowledge, making it less effective in handling both MSL and MTL. In this paper, we propose a novel Automatic Expert Selection framework for Multi-scenario and Multi-task search, named AESM^{2}. AESM^{2} integrates both MSL and MTL into a unified framework with an automatic structure learning. Specifically, AESM^{2} stacks multi-task layers over multi-scenario layers. This hierarchical design enables us to flexibly establish intrinsic connections between different scenarios, and at the same time also supports high-level feature extraction for different tasks. At each multi-scenario/multi-task layer, a novel expert selection algorithm is proposed to automatically identify scenario-/task-specific and shared experts for each input. Experiments over two real-world large-scale datasets demonstrate the effectiveness of AESM^{2} over a battery of strong baselines. Online A/B test also shows substantial performance gain on multiple metrics. Currently, AESM^{2} has been deployed online for serving major traffic.