From Agnostic to Specific: Latent Preference Diffusion for Multi-Behavior Sequential Recommendation Conference: Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1; August 09–13, 2026; Jeju Island, Republic of Korea Proceedings of the 32nd ACM SIGKDD Conference o
Ruochen Yang, Xiaodong Li, Jiawei Sheng, Jiangxia Cao, Xinkui Lin, Shen Wang, Shuang Yang, Zhaojie Liu, Tingwen Liu · 2026
Read From Agnostic to Specific: Latent Preference Diffusion for Multi-Behavior Sequential Recommendation Conference: Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1; August 09–13, 2026; Jeju Island, Republic of Korea Proceedings of the 32nd ACM SIGKDD Conference o by Ruochen Yang, Xiaodong Li, Jiawei Sheng, Jiangxia Cao, Xinkui Lin, Shen Wang, Shuang Yang, Zhaojie Liu, Tingwen Liu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-23132.txt.
Description
Ruochen Yang and eight co-authors guide taste spread from conduct-open to conduct-set. The task learns mixed links in user sequences to capture tastes under target conduct. Earlier methods map auxiliary conducts to the target conduct one-way, while recent conduct-set methods ignore the user's hidden taste behind choices.
Meanwhile, the lopsided link between items and conducts makes taste scoring unsuitable to capture doubt from low-chaos conducts to high-chaos items. It misses lean and varied picks. To address these challenges, the team proposes FatsMB, a spread model steering taste making from conduct-open to conduct-set in hidden spaces, enabling varied and true picks. A Multi-Conduct AutoEncoder builds a single hidden taste space, easing links across conducts. Next, target taste transfer boosts with telling priors. Multi-Condition Guided Layer Normalization is added for the cleaning. Wide tests show the model's value.
Catalog record
| CREATOR | Ruochen Yang, Xiaodong Li, Jiawei Sheng, Jiangxia Cao, Xinkui Lin, Shen Wang, Shuang Yang, Zhaojie Liu, Tingwen Liu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-23132