Distillation-based Scenario-Adaptive Mixture-of-Experts for the Matching Stage of Multi-scenario Recommendation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Ruibing, Guo, Shuhan, Du, Haotong, Yao, Quanming
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917185678475264
author Wang, Ruibing
Guo, Shuhan
Du, Haotong
Yao, Quanming
author_facet Wang, Ruibing
Guo, Shuhan
Du, Haotong
Yao, Quanming
contents Multi-scenario recommendation is pivotal for optimizing user experience across diverse contexts. While Multi-gate Mixture-of-Experts (MMOE) thrives in ranking, its transfer to the matching stage is hindered by the blind optimization inherent to independent two-tower architectures and the parameter dominance of head scenarios. To address these structural and distributional bottlenecks, we propose Distillation-based Scenario-Adaptive Mixture-of-Experts (DSMOE). Specially, we devise a Scenario-Adaptive Projection (SAP) module to generate lightweight, context-specific parameters, effectively preventing expert collapse in long-tail scenarios. Concurrently, we introduce a cross-architecture knowledge distillation framework, where an interaction-aware teacher guides the two-tower student to capture complex matching patterns. Extensive experiments on real-world datasets demonstrate DSMOE's superiority, particularly in significantly improving retrieval quality for under-represented, data-sparse scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2601_02368
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Distillation-based Scenario-Adaptive Mixture-of-Experts for the Matching Stage of Multi-scenario Recommendation
Wang, Ruibing
Guo, Shuhan
Du, Haotong
Yao, Quanming
Information Retrieval
Artificial Intelligence
Multi-scenario recommendation is pivotal for optimizing user experience across diverse contexts. While Multi-gate Mixture-of-Experts (MMOE) thrives in ranking, its transfer to the matching stage is hindered by the blind optimization inherent to independent two-tower architectures and the parameter dominance of head scenarios. To address these structural and distributional bottlenecks, we propose Distillation-based Scenario-Adaptive Mixture-of-Experts (DSMOE). Specially, we devise a Scenario-Adaptive Projection (SAP) module to generate lightweight, context-specific parameters, effectively preventing expert collapse in long-tail scenarios. Concurrently, we introduce a cross-architecture knowledge distillation framework, where an interaction-aware teacher guides the two-tower student to capture complex matching patterns. Extensive experiments on real-world datasets demonstrate DSMOE's superiority, particularly in significantly improving retrieval quality for under-represented, data-sparse scenarios.
title Distillation-based Scenario-Adaptive Mixture-of-Experts for the Matching Stage of Multi-scenario Recommendation
topic Information Retrieval
Artificial Intelligence
url https://arxiv.org/abs/2601.02368