Precision marketing recommendations face significant challenges due to entanglement of sharing in shared-account marketing. To disentangle shared user behaviors within strict privacy policies, proposes the Hierarchical Self-Attentive Sequential Recommendation (HierSASRec) model. HierSASRec ploys user disambiguation learning to dynamically identify individual users within shared accounts, two-level representations for account-item and user-item interactions. By integrating the time intervals interactions into the density-based spatial clustering of applications with noise (DBSCAN) method, namely, time-aware DBSCAN method, HierSASRec automatically extracts user-level sequences beyond fixed user enhancing the identification of similar preferences and close interactions. Through a self-attention mechanism, HierSASRec combines hierarchical interaction information to optimize marketing precision. Additionally, random user switch mechanism is devised to mitigate noise from long-term sequences, and focuses immediate marketing decisions for the current user. Experimental validation within real-world underscores the superiority of HierSASRec over state-of-the-art baselines, affirming its practical efficacy enhancing marketing precision. The code is available at: https://github.com/muyunping123/HierSASRec.
