ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914223057010688 |
|---|---|
| author | Tang, Jiakai Wang, Chuan Yang, Gaoming Wu, Han Yu, Jiahao Wu, Jian Hu, Jianwu Zheng, Junjun Li, Longbin Xiao, Shuwen Kong, Xiangheng Yang, Yeqiu Jiang, Yuning Nurlanbek, Ahjol Cao, Binbin Zheng, Bo Zhu, Fangmei Zhou, Gaoming Yi, Huimin Chu, Huiping Huang, Jin Shan, Jinzhe Cui, Kenan Li, Longbin Zhou, Silu Chen, Wen Ming, Xia Gao, Xiang Yao, Xin Wen, Xingyu Zhang, Yan Hu, Yiwen Wang, Yulin Bao, Ziheng Wu, Zongyuan |
| author_facet | Tang, Jiakai Wang, Chuan Yang, Gaoming Wu, Han Yu, Jiahao Wu, Jian Hu, Jianwu Zheng, Junjun Li, Longbin Xiao, Shuwen Kong, Xiangheng Yang, Yeqiu Jiang, Yuning Nurlanbek, Ahjol Cao, Binbin Zheng, Bo Zhu, Fangmei Zhou, Gaoming Yi, Huimin Chu, Huiping Huang, Jin Shan, Jinzhe Cui, Kenan Li, Longbin Zhou, Silu Chen, Wen Ming, Xia Gao, Xiang Yao, Xin Wen, Xingyu Zhang, Yan Hu, Yiwen Wang, Yulin Bao, Ziheng Wu, Zongyuan |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_21257 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling Tang, Jiakai Wang, Chuan Yang, Gaoming Wu, Han Yu, Jiahao Wu, Jian Hu, Jianwu Zheng, Junjun Li, Longbin Xiao, Shuwen Kong, Xiangheng Yang, Yeqiu Jiang, Yuning Nurlanbek, Ahjol Cao, Binbin Zheng, Bo Zhu, Fangmei Zhou, Gaoming Yi, Huimin Chu, Huiping Huang, Jin Shan, Jinzhe Cui, Kenan Li, Longbin Zhou, Silu Chen, Wen Ming, Xia Gao, Xiang Yao, Xin Wen, Xingyu Zhang, Yan Hu, Yiwen Wang, Yulin Bao, Ziheng Wu, Zongyuan Information Retrieval Computation and Language 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. |
| title | ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling |
| topic | Information Retrieval Computation and Language |
| url | https://arxiv.org/abs/2512.21257 |