SIGMA: A Semantic-Grounded Instruction-Driven Generative Multi-Task Recommender at AliExpress

TL;DR

This work presents SIGMA, a Semantic-Grounded Instruction-Driven Generative Multi-Task Recommender, a Semantic-Grounded Instruction-Driven Generative Multi-Task Recommender that first ground item entities in general semantics via a unified latent space capturing both semantic and collaborative relations.

Abstract

With the rapid evolution of Large Language Models, generative recommendation is gradually reshaping the paradigm of recommender systems. However, most existing methods are still confined to the interaction-driven next-item prediction paradigm, failing to rapidly adapt to evolving trends or address diverse recommendation tasks along with business-specific requirements in real-world scenarios. To this end, we present SIGMA, a Semantic-Grounded Instruction-Driven Generative Multi-Task Recommender at AliExpress. Specifically, we first ground item entities in general semantics via a unified latent space capturing both semantic and collaborative relations. Building upon this, we develop a hybrid item tokenization method for precise modeling and efficient generation. Moreover, we construct a large-scale multi-task SFT dataset to empower SIGMA to fulfill various recommendation demands via instruction-following. Finally, we design a three-step item generation procedure integrated with an adaptive probabilistic fusion mechanism to calibrate the output distributions based on task-specific requirements for recommendation accuracy and diversity. Extensive offline experiments and online A/B tests demonstrate the effectiveness of SIGMA.

Figures (3)

• Figure 1: The overall framework of SIGMA.
• Figure 2: Performance variations of different methods on each task across with model scaling.
• Figure 3: The online serving architecture for SIGMA.