ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling

TL;DR

ReaSeq shifts industrial ranking from pure log-driven co-occurrence to world-knowledge–driven reasoning by combining explicit, multi-agent Chain-of-Thought item knowledge with latent diffusion-based behavior reasoning. It introduces a structured knowledge system to produce semantically enriched item embeddings and two complementary sequential modeling paths (retrieval-based and compression-based) to leverage this knowledge in ranking. Separately, Generative Behavior Reasoning (GBR) locates beyond-log discontinuities and generates plausible augmentations via a DLLM framework, addressing blind spots in user interests. In deployment on Taobao, ReaSeq delivers consistent gains in IPV, CTR, Orders, and GMV, demonstrating the practical value of world-knowledge–enabled reasoning for scalable, industrial recommender systems.

Abstract

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.

Figures (7)

• Figure 1: Illustration of different sequential modeling paradigms, i.e. traditional log-driven and the proposed ReaSeq. Our ReaSeq fully unleash world knowledge of LLMs through reasoning techniques, which not only enriches representations to mitigate the knowledge poverty of IDs but also utilize behavior generation to expand model's perception of user interests beyond the log.
• Figure 2: Architectural overview of the proposed ReaSeq framework. ReaSeq is composed of two synergistic parts: (1) Knowledge System: This offline module constructs two core assets. Reasoning-Enhanced Representation employs a multi-agent system to generate semantic embeddings from user demand and product attributes. Generative Behavior Reasoning uses a DLLM to locate and reconstruct plausible beyond-log user behaviors. (2) Application: This online module applies the knowledge assets to enhance sequential modeling. It supports two paradigms: a Retrieval-Based Model for GSU-ESU architectures and a Compression-Based Model that uses target-aware interest extraction for long-sequence modeling.
• Figure 3: The workflow of our Multi-Agent Knowledge Reasoning framework. (1) Categorical Information Extraction Agents first distill a dual-perspective taxonomy (User-centric Demand and Product-centric Attribute) from category-wide user queries. (2) Item-Specific Knowledge Generation Agents are then prompted with this taxonomy to reason over an item's raw content (Item CPV), systematically populating the defined dimensions with specific values (e.g., 'Daily Commute').
• Figure 4: Overall architecture of the ranking model, processing user sequences through two complementary paths. (1) The Retrieval-Based Modeling employs a Top-K Retrieval Service to create a target-relevant subsequence, followed by an efficient batch-gathered attention mechanism for online scoring. (2) The Compression-Based Modeling uses learnable interest anchors to distill the entire long sequence into a compact representation, which is then queried by the target item. The resulting vectors from both paths are combined with other features for the final pCTR prediction.
• Figure 5: The two-stage process for generative behavior reasoning. (1) Where: Beyond-Log Behavior Location. A two-step filtering pipeline pinpoints potential discontinuities, starting with a rule-based filter (e.g., temporal or category gaps) followed by a model-based filter that uses a standard recommender to identify low-probability transitions. (2) What: Bidirectional Implicit Behavior Reasoning. A DLLM-based model processes the sequence. During training, it learns to reconstruct masked in-log behaviors ([M]); during inference, it performs offline generation to fill in the identified beyond-log discontinuities ([F]).