OneMall: One Architecture, More Scenarios -- End-to-End Generative Recommender Family at Kuaishou E-Commerce

2026-01-29

AI Summary

This work addresses the challenge of fragmented recommendation systems across multiple e-commerce scenarios—product cards, short videos, and live streaming—which hinder unified modeling. To this end, we propose OneMall, an end-to-end generative recommendation framework that, for the first time in e-commerce, enables cross-scenario unified generative recommendations. Key innovations include a semantic-aware e-commerce tokenizer, Query-Former for long-sequence compression, cross-attention-based multi-behavior fusion, sparse MoE-based autoregressive generation, and reinforcement learning-driven joint optimization of retrieval and ranking. Extensive experiments demonstrate significant performance gains across all scenarios: a 13.01% increase in GMV for product cards, a 15.32% rise in order volume for short videos, and a 2.78% improvement in live-streaming orders. OneMall has been deployed in an industrial system serving over 400 million daily active users.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Search and Retrieval-Augmented AI: Ad search and search for Web retail

Abstract

In the wave of generative recommendation, we present OneMall, an end-to-end generative recommendation framework tailored for e-commerce services at Kuaishou. Our OneMall systematically unifies the e-commerce's multiple item distribution scenarios, such as Product-card, short-video and live-streaming. Specifically, it comprises three key components, aligning the entire model training pipeline to the LLM's pre-training/post-training: (1) E-commerce Semantic Tokenizer: we provide a tokenizer solution that captures both real-world semantics and business-specific item relations across different scenarios; (2) Transformer-based Architecture: we largely utilize Transformer as our model backbone, e.g., employing Query-Former for long sequence compression, Cross-Attention for multi-behavior sequence fusion, and Sparse MoE for scalable auto-regressive generation; (3) Reinforcement Learning Pipeline: we further connect retrieval and ranking models via RL, enabling the ranking model to serve as a reward signal for end-to-end policy retrieval model optimization. Extensive experiments demonstrate that OneMall achieves consistent improvements across all e-commerce scenarios: +13.01\% GMV in product-card, +15.32\% Orders in Short-Video, and +2.78\% Orders in Live-Streaming. OneMall has been deployed, serving over 400 million daily active users at Kuaishou.

Problem

• generative recommendation
• e-commerce
• multi-scenario
• recommendation system
• item distribution

Innovation

• Generative Recommendation
• End-to-End Framework
• Transformer Architecture
• Reinforcement Learning
• E-commerce Semantic Tokenizer