RoTE: Coarse-to-Fine Multi-Level Rotary Time Embedding for Sequential Recommendation

TL;DR

RoTE introduces a multi-level temporal embedding module that explicitly models time span information in sequential recommendation, improving the capture of user interest dynamics.

Contribution

It proposes a novel coarse-to-fine multi-level temporal embedding method that enhances existing Transformer-based recommendation models by explicitly modeling temporal spans.

Findings

Achieves up to 20.11% improvement in NDCG@5 on benchmark datasets.

Effectively captures heterogeneous temporal patterns in user interactions.

Can be integrated into existing models without modifying their backbone architectures.

Abstract

Sequential recommendation models have been widely adopted for modeling user behavior. Existing approaches typically construct user interaction sequences by sorting items according to timestamps and then model user preferences from historical behaviors. While effective, such a process only considers the order of temporal information but overlooks the actual time spans between interactions, resulting in a coarse representation of users' temporal dynamics and limiting the model's ability to capture long-term and short-term interest evolution. To address this limitation, we propose RoTE, a novel multi-level temporal embedding module that explicitly models time span information in sequential recommendation. RoTE decomposes each interaction timestamp into multiple temporal granularities, ranging from coarse to fine, and incorporates the resulting temporal representations into item embeddings. This design enables models to capture heterogeneous temporal patterns and better perceive temporal distances among user interactions during sequence modeling. RoTE is a lightweight, plug-and-play module that can be seamlessly integrated into existing Transformer-based sequential recommendation models without modifying their backbone architectures. We apply RoTE to several representative models and conduct extensive experiments on three public benchmarks. Experimental results demonstrate that RoTE consistently enhances the corresponding backbone models, achieving up to a 20.11% improvement in NDCG@5, which confirms the effectiveness and generality of the proposed approach. Our code is available at https://github.com/XiaoLongtaoo/RoTE.