UxSID: Semantic-Aware User Interests Modeling for Ultra-Long Sequence

TL;DR

UxSID introduces a semantic-aware framework for ultra-long user interest modeling that balances efficiency and effectiveness, achieving state-of-the-art results in large-scale advertising.

Contribution

It proposes a novel semantic-group shared interest memory using Semantic IDs and dual-level attention, offering a new approach beyond item-specific search or item-agnostic compression.

Findings

Achieves state-of-the-art performance in user interest modeling.

Realizes a 0.337% revenue lift in large-scale advertising A/B testing.

Balances computational efficiency with semantic awareness.

Abstract

Modeling ultra-long user sequences involves a difficult trade-off between efficiency and effectiveness. While current paradigms rely on either item-specific search or item-agnostic compression, we propose UxSID, a framework exploring a third path: semantic-group shared interest memory. By utilizing Semantic IDs (SIDs) and a dual-level attention strategy, UxSID captures target-aware preferences without the heavy cost of item-specific models. This end-to-end architecture balances computational parsimony with semantic awareness, achieving state-of-the-art performance and a 0.337% revenue lift in large-scale advertising A/B test.