DisenReason: Behavior Disentanglement and Latent Reasoning for Shared-Account Sequential Recommendation

2026-03-04

AI Summary

This work addresses the challenge of accurately inferring the number of latent users and their behavioral patterns in shared account scenarios, a task where existing methods are constrained by assumptions of a fixed user count. To overcome this limitation, the authors propose DisenReason, a two-stage inference framework that first constructs a unified collective account representation through frequency-domain behavior disentanglement and then leverages this representation as a pivot to dynamically infer both the true number of users and their individual interests. Innovatively, the problem of user count estimation is reformulated as an intermediate embedding generation task, enabling adaptive modeling of multi-user structures. Extensive experiments on four benchmark datasets demonstrate significant performance gains over state-of-the-art methods, with improvements of up to 12.56% in MRR@5 and 6.06% in Recall@20.

Technology Category

• Knowledge Representation and Reasoning: Preferences
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: The sharing economy

Abstract

Shared-account usage is common on streaming and e-commerce platforms, where multiple users share one account. Existing shared-account sequential recommendation (SSR) methods often assume a fixed number of latent users per account, limiting their ability to adapt to diverse sharing patterns and reducing recommendation accuracy. Recent latent reasoning technique applied in sequential recommendation (SR) generate intermediate embeddings from the user embedding (e.g, last item embedding) to uncover users' potential interests, which inspires us to treat the problem of inferring the number of latent users as generating a series of intermediate embeddings, shifting from inferring preferences behind user to inferring the users behind account. However, the last item cannot be directly used for reasoning in SSR, as it can only represent the behavior of the most recent latent user, rather than the collective behavior of the entire account. To address this, we propose DisenReason, a two-stage reasoning method tailored to SSR. DisenReason combines behavior disentanglement stage from frequency-domain perspective to create a collective and unified account behavior representation, which serves as a pivot for latent user reasoning stage to infer the number of users behind the account. Experiments on four benchmark datasets show that DisenReason consistently outperforms all state-of-the-art baselines across four benchmark datasets, achieving relative improvements of up to 12.56\% in MRR@5 and 6.06\% in Recall@20.

Problem

• shared-account sequential recommendation
• latent user inference
• behavior disentanglement
• sequential recommendation
• collective behavior representation

Innovation

• behavior disentanglement
• latent reasoning
• shared-account sequential recommendation
• frequency-domain representation
• dynamic user inference