A Generative Re-ranking Model for List-level Multi-objective Optimization at Taobao

TL;DR

This work tackles list-level multi-objective optimization in e-commerce re-ranking, addressing the limitations of fixed-weight, item-level approaches. It introduces SORT-Gen, a generative re-ranking framework consisting of a Sequential Ordered Regression Transformer to model real-time user intent and a Mask-Driven Fast Generation Algorithm for efficient, one-pass list generation. The model uses an ordered regression loss to estimate multi-objective values for variable-length sub-lists and integrates an MMR-based mechanism to balance accuracy and diversity, achieving significant online gains in CLICK and GMV with low latency. The approach is deployed in Taobao across multiple scenarios, demonstrating practical impact on large-scale recommender systems.

Abstract

E-commerce recommendation systems aim to generate ordered lists of items for customers, optimizing multiple business objectives, such as clicks, conversions and Gross Merchandise Volume (GMV). Traditional multi-objective optimization methods like formulas or Learning-to-rank (LTR) models take effect at item-level, neglecting dynamic user intent and contextual item interactions. List-level multi-objective optimization in the re-ranking stage can overcome this limitation, but most current re-ranking models focus more on accuracy improvement with context. In addition, re-ranking is faced with the challenges of time complexity and diversity. In light of this, we propose a novel end-to-end generative re-ranking model named Sequential Ordered Regression Transformer-Generator (SORT-Gen) for the less-studied list-level multi-objective optimization problem. Specifically, SORT-Gen is divided into two parts: 1)Sequential Ordered Regression Transformer innovatively uses Transformer and ordered regression to accurately estimate multi-objective values for variable-length sub-lists. 2)Mask-Driven Fast Generation Algorithm combines multi-objective candidate queues, efficient item selection and diversity mechanism into model inference, providing a fast online list generation method. Comprehensive online experiments demonstrate that SORT-Gen brings +4.13% CLCK and +8.10% GMV for Baiyibutie, a notable Mini-app of Taobao. Currently, SORT-Gen has been successfully deployed in multiple scenarios of Taobao App, serving for a vast number of users.

Figures (3)

• Figure 1: An overview of the SORT-Gen framework.
• Figure 2: Online real-world cases where SORT-Gen works.
• Figure 3: Multi-objective value based on real logs.