Search papers, labs, and topics across Lattice.
The SPARC framework enhances generative recommendation systems by modeling sequential dependencies of user behavior attributes to create context-aware representations. By routing and compressing these representations, SPARC effectively retains relevant information while maintaining a fixed input length, thus avoiding the pitfalls of static attribute assignments. Experimental results on Taobao and Amazon datasets show that SPARC significantly outperforms both conventional and generative baselines, highlighting the importance of context in user-history representation.
Context-aware routing and compression in SPARC leads to superior generative recommendations by preserving crucial information without inflating input size.
Generative recommendation tokenizes items as discrete Semantic IDs (SIDs) and autoregressively generates target items from users'historical SID sequences. Although existing SIDs incorporate multimodal and structured information, they are typically statically assigned and independent of the current interaction context. In industrial scenarios, each behavior also contains heterogeneous attributes, such as category, brand, price, behavior type, and timestamp. Fully expanding these features greatly increases the input length, while directly compressing them into a single representation may prematurely discard context-relevant information. We propose \textbf{SPARC}, \uline{\textbf{S}}equence-aware \uline{\textbf{P}}rogressive \uline{\textbf{A}}ttribute \uline{\textbf{R}}outing and \uline{\textbf{C}}ompression Framework for Generative recommendation. SPARC first models the sequential dependencies of each field type to obtain context-aware field representations. It then routes the original, contextual, and identity representations of different fields into multiple slots to preserve complementary information under a fixed capacity. Finally, lightweight cross-item interaction integrates the intermediate tokens and compresses each historical item into a single token. Following the principle of contextualizing before compression, SPARC enriches user-history representations without increasing the input length of the generative backbone. Experiments on industrial Taobao and public Amazon datasets demonstrate that SPARC outperforms strong conventional and generative baselines. Further comparisons with static compression variants show that the improvement of SPARC comes from context-conditioned information retention rather than merely increasing the expressiveness of the compression module.
DAMO