Large language models (LLMs) are increasingly adopted as the backbone of recommender systems. However, user–item interactions in real-world scenarios are non-stationary, making preference drift over time inevitable. Existing model update strategies mainly rely on global fine-tuning or pointwise editing, but they face two fundamental challenges: (i) imbalanced update granularity, where global updates perturb behaviors unrelated to the target while pointwise edits fail to capture broader preference shifts; (ii) unstable incremental updates, where repeated edits interfere with prior adaptations, leading to catastrophic forgetting and inconsistent recommendations. To address these issues, we propose Region-Aware Incremental Editing (RAIE), a plug-in framework that freezes the backbone model and performs region-level updates. RAIE first constructs semantically coherent preference regions via spherical k-means in the representation space. It then assigns incoming sequences to regions via confidence-aware gating and performs three localized edit operations-Update, Expand, and Add-to dynamically revise the affected region. Each region is equipped with a dedicated Low-Rank Adaptation (LoRA) module, which is trained only on the region's updated data. During inference, RAIE routes each user sequence to its corresponding region and activates the region-specific adapter for prediction. Experiments on two benchmark datasets under a time-sliced protocol that segments data into Set-up (S), Finetune (F), and Test (T) show that RAIE significantly outperforms state-of-the-art baselines while effectively mitigating forgetting. These results demonstrate that region-aware editing offers an accurate and scalable mechanism for continual adaptation in dynamic recommendation scenarios.

Previous user-item interaction graphs have typically focused on simple interaction between users and items, failing to identify the important effects of user's intents in the interaction. While recent studies have ventured into exploring intent relationships between users and items for modeling, they predominantly emphasize user preferences manifesting in the interaction, overlooking knowledge-driven insight, thereby limiting the interpretability of intent. In this paper, we utilize the rich interpretable knowledge information in the knowledge graph to design a novel dual-level intents modeling framework called DIM. DIM aims to mine user's true intents, which usually include user popularity preference and personalized preference. Therefore, we extract both the popular and personalized user preferences from attribute tuples within the knowledge graph at the global and local levels, respectively. Experimental results on three datasets demonstrate the superiority of DIM over various state-of-the-art approaches.

Recommendation systems have shifted from personalization for individual users to consensus for groups as a result of people’s growing tendency to join groups to participate in various everyday activities, like family meals and workplace reunions. This is because social networks have made it easier for people to participate in these kinds of events. Group recommendation is the process of suggesting items to groups. To derive group preferences, the majority of current approaches combine the individual preferences of group members utilizing heuristic or attention mechanism-based techniques. These approaches, however, have three issues. First, these approaches ignore the complex high-order interactions that occur both inside and outside of groups, just modeling the preferences of individual groups of users. Second, a group’s ultimate decision is not always determined by the members’ preferences. Nevertheless, current approaches are not adequate to represent such preferences across groups. Last, data sparsity affects group recommendations due to the sparsity of group–item interactions. To overcome the aforementioned constraints, we propose employing hypergraph convolutional networks for group recommendation. Specifically, our design aims to achieve excellent group preferences by establishing a high-order preference extraction view represented by the hypergraph, a consistent preference extraction view represented by the overlap graph, and a conventional preference extraction view represented by the bipartite graph. The linkages between the three various views are then established by using cross-view contrastive learning, and the information between different views can be complementary, thereby improving each other. Comprehensive experiments on three publicly available datasets show that our method performs better than the state-of-the-art baseline.