ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling
This addresses the problem of brittle and constrained performance in industrial recommender systems for platforms like Taobao, representing a novel method for a known bottleneck rather than an incremental improvement.
The paper tackles the limitations of log-driven recommender systems, such as knowledge poverty and blindness to beyond-log interests, by introducing ReaSeq, a reasoning-enhanced framework that leverages world knowledge from Large Language Models, achieving gains of over 6.0% in IPV and CTR, over 2.9% in Orders, and over 2.5% in GMV on Taobao's ranking system.
Industrial recommender systems face two fundamental limitations under the log-driven paradigm: (1) knowledge poverty in ID-based item representations that causes brittle interest modeling under data sparsity, and (2) systemic blindness to beyond-log user interests that constrains model performance within platform boundaries. These limitations stem from an over-reliance on shallow interaction statistics and close-looped feedback while neglecting the rich world knowledge about product semantics and cross-domain behavioral patterns that Large Language Models have learned from vast corpora. To address these challenges, we introduce ReaSeq, a reasoning-enhanced framework that leverages world knowledge in Large Language Models to address both limitations through explicit and implicit reasoning. Specifically, ReaSeq employs explicit Chain-of-Thought reasoning via multi-agent collaboration to distill structured product knowledge into semantically enriched item representations, and latent reasoning via Diffusion Large Language Models to infer plausible beyond-log behaviors. Deployed on Taobao's ranking system serving hundreds of millions of users, ReaSeq achieves substantial gains: >6.0% in IPV and CTR, >2.9% in Orders, and >2.5% in GMV, validating the effectiveness of world-knowledge-enhanced reasoning over purely log-driven approaches.