Decoupled Temporal Encoding for Generative Recommendation
2026-08-17 • Information Retrieval
Information RetrievalArtificial Intelligence
AI summaryⓘ
The authors explain that current recommendation systems, especially those based on Transformers, usually consider the order of items but often miss important timing details, like when interactions happen or special events influencing user behavior. They introduce Decoupled Temporal Encoding (DTE), a new method that separates broad timing patterns from the order of items to better capture complex time effects such as daily routines and promotions. Their approach uses two parts: one to add general time info personalized to users, and another that emphasizes order only when interactions happen closely in time. This method is efficient and easy to add to existing recommendation systems.
Positional encodingTransformer modelsGenerative recommendationTemporal dynamicsAutoregressive sequencesPersonalized modelingTime-gated encodingMicro-sequential orderMacro-temporal patternsParameter efficiency
Authors
Pengfei Jia, Jingjian Wang, Jingmao Li, Ge Zhang, Feng Shi
Abstract
Positional encoding is a fundamental component of Transformer-based generative recommendation models, where user histories are modeled as autoregressive item sequences. Most positional encoding methods are inherited from natural language processing and mainly represent discrete item order. However, recommendation sequences go beyond ordered lists, as timestamps and temporal effects also shape item relations. Our work is motivated by a real-world food delivery and instant retail recommendation system, where user behavior exhibits multi-level temporal regularities, including recency effects, meal-time peaks, weekday-weekend shifts, and promotion-driven traffic bursts. Existing methods partially address this issue through timestamp features, interval embeddings, decay functions, or attention biases, but they usually inject heterogeneous temporal signals through a unified representation or a single modeling pathway, making it difficult to distinguish broad temporal dynamics from local order cues. To address this limitation, we propose Decoupled Temporal Encoding, a lightweight framework for generative recommendation. DTE separates temporal dynamics from order information through two complementary modules: a personalized macro-temporal module that injects compact temporal primitives into item embeddings, and a time-gated micro-sequential module that introduces relative-order bias only when interactions are temporally dense. DTE is also parameter-efficient and deployment-friendly, allowing easy integration into existing systems.