Despite growing efforts to mitigate unfairness in recommender systems, existing fairness-aware methods typically fix the fairness requirement at training time and provide limited post-training flexibility. However, in real-world scenarios, diverse stakeholders may demand differing fairness requirements over time, so retraining for different fairness requirements becomes prohibitive. To address this limitation, we propose Cofair, a single-train framework that enables post-training fairness control in recommendation. Specifically, Cofair introduces a shared representation layer with fairness-conditioned adapter modules to produce user embeddings specialized for varied fairness levels, along with a user-level regularization term that guarantees user-wise monotonic fairness improvements across these levels. We theoretically establish that the adversarial objective of Cofair upper bounds demographic parity and the regularization term enforces progressive fairness at user level. Comprehensive experiments on multiple datasets and backbone models demonstrate that our framework provides dynamic fairness at different levels, delivering comparable or better fairness-accuracy curves than state-of-the-art baselines, without the need to retrain for each new fairness requirement. Our code is publicly available at https://github.com/weixinchen98/Cofair.

Cross-domain recommendation (CDR) offers an effective strategy for improving recommendation quality in a target domain by leveraging auxiliary signals from source domains. Nonetheless, emerging evidence shows that CDR can inadvertently heighten group-level unfairness. In this work, we conduct a comprehensive theoretical and empirical analysis to uncover why these fairness issues arise. Specifically, we identify two key challenges: (i) Cross-Domain Disparity Transfer, wherein existing group-level disparities in the source domain are systematically propagated to the target domain; and (ii) Unfairness from Cross-Domain Information Gain, where the benefits derived from cross-domain knowledge are unevenly allocated among distinct groups. To address these two challenges, we propose a Cross-Domain Fairness Augmentation (CDFA) framework composed of two key components. Firstly, it mitigates cross-domain disparity transfer by adaptively integrating unlabeled data to equilibrate the informativeness of training signals across groups. Secondly, it redistributes cross-domain information gains via an information-theoretic approach to ensure equitable benefit allocation across groups. Extensive experiments on multiple datasets and baselines demonstrate that our framework significantly reduces unfairness in CDR without sacrificing overall recommendation performance, while even enhancing it.

Representation learning is essential for deep-neural-network-based recommender systems to capture user preferences and item features within fixed-dimensional user and item vectors. Unlike existing representation learning methods that either treat each user preference and item feature uniformly or categorize them into discrete clusters, we argue that in the real world, user preferences and item features are naturally expressed and organized in a hierarchical manner, leading to a new direction for representation learning. In this article, we introduce a novel matryoshka representation learning method for recommendation (MRL4Rec) , by which we restructure user and item vectors into matryoshka representations with nested vector spaces to explicitly represent user preferences and item features at different hierarchical layers. We theoretically establish that training with the same triplets for each sliced vector cannot guarantee representation learning with hierarchical structures. Subsequently, we propose the layer- and hardness-adaptive negative sampling (LHANS) mechanism to construct training triplets, which further ensures the soundness of learned matryoshka representations in capturing hierarchical user preferences and item features. The experiments demonstrate that MRL4Rec can consistently and substantially outperform a number of state-of-the-art competitors on several real-life datasets. Our code is publicly available at https://github.com/Riwei-HEU/MRL .

Large language model (LLM) based recommendation agents personalize what they know through evolving per-user semantic memory, yet how they reason remains a universal, static system prompt shared identically across all users. This asymmetry is a fundamental bottleneck: when a recommendation fails, the agent updates its memory of user preferences but never interrogates the decision logic that produced the failure, leaving its reasoning process structurally unchanged regardless of how many mistakes it accumulates. To address this bottleneck, we propose SAGER (Self-Evolving Agent for Personalized Recommendation), the first recommendation agent framework in which each user is equipped with a dedicated policy skill, a structured natural-language document encoding personalized decision principles that evolves continuously through interaction. SAGER introduces a two-representation skill architecture that decouples a rich evolution substrate from a minimal inference-time injection, an incremental contrastive chain-of-thought engine that diagnoses reasoning flaws by contrasting accepted against unchosen items while preserving accumulated priors, and skill-augmented listwise reasoning that creates fine-grained decision boundaries where the evolved skill provides genuine discriminative value. Experiments on four public benchmarks demonstrate that SAGER achieves state-of-the-art performance, with gains orthogonal to memory accumulation, confirming that personalizing the reasoning process itself is a qualitatively distinct source of recommendation improvement.

Offline evaluation is the predominant method for scientific research in recommender systems, enabling the comparison of alternative recommendation approaches using pre-collected datasets and computational metrics without involving human participants. However, recent reproducibility studies reveal that many offline evaluations in the literature lack scientific rigor or adopt research practices that cast doubt on the validity of their findings. A particularly common and ultimately catastrophic flaw is the comparison of newly proposed machine learning models against untuned or poorly tuned baseline models. Combined with limited reproducibility, such practices raise serious concerns about the true progress achieved by increasingly complex recommendation algorithms. In this editorial, we argue for stronger methodological standards and summarize essential guidance and best practices for conducting rigorous offline evaluations of recommender systems. Accordingly, ACM Transactions on Recommender Systems will place increased emphasis on methodological rigor in all future submissions, with particular priority given to work that provides comprehensive reproducibility materials and clearly documents the tuning procedures used for baseline models.

Conversational recommender systems (CRS) allow users to express preferences and provide feedback through natural language. With the emergence of ChatGPT, there is growing interest in leveraging its capabilities to enhance user engagement and improve recommendation quality. Research suggests that well-crafted prompts are essential for ChatGPT to accurately interpret tasks and generate high-quality responses. Therefore, we define Prompt Guidance (PG) as offering an example of crafting queries (prompts) that clearly specify the context, constraints, and procedures for the recommendations users may seek. Offering Prompt Guidance (PG) to novices can improve the overall quality of conversations. Meanwhile, in the field of recommender systems, Recommendation Domain (RD) can influence user behavior and perception. This study explores how PG and RD interact to affect user experience (UX) in ChatGPT-based recommendations. To investigate this, we conducted an empirical study with 100 participants, analyzing dialogue data to uncover user intents and system actions. Using a within-subject (e.g., book vs. job recommendations) and between-subject (providing prompts vs. none) design, we evaluated UX under different experimental conditions. Results show that PG enhances explainability, adaptability, ease of use, and transparency. In terms of the RD factor, the system promotes higher user engagement, greater novelty, and a stronger intention to try suggested items in book recommendations compared to job recommendations. Overall, ChatGPT-based recommendations not only induce novel dialogue behaviors but also improve UX in multiple aspects, such as usefulness and alignment with user preferences. These findings provide valuable insights into designing CRS that leverage ChatGPT’s potential to deliver effective and engaging user experiences.

Predicting a user's next search query from recent interaction behaviors is a critical problem in modern e-commerce systems, particularly in scenarios where user intent evolves rapidly. Large Language Models (LLMs) offer strong semantic reasoning capabilities and have recently been adopted to enhance training data construction for next-query prediction. However, due to resource constraints on mobile devices, existing applications are deployed on cloud servers, resulting in high inference costs. In this paper, we propose RecGPT-Mobile, a framework that designs a lightweight LLM-based intent understanding agent to improve recommendation quality in mobile e-commerce scenarios. By deploying LLMs directly on mobile devices, our approach can capture evolving interests of users more quickly and adjust the recommendation results in real time. Extensive offline analyses and online experiments demonstrate that our method significantly improves the accuracy of recommendation results, laying a practical path for LLM deployment in production-scale recommendation systems on mobile devices, as well as a scalable solution for integrating LLMs into real-world next-query prediction systems.

The evolution of recommender systems has shifted from traditional collaborative filtering to LLM-based agentic systems, which rely on semantic user and item memories to make predictions. However, existing agents maintain these memories in isolation. This overlooks crucial collaborative signals, such as user-item co-engagements and peer relationships across the community, which significantly limits their ability to uncover hidden preferences and accurately infer user needs, particularly for data-sparse users. To bridge this gap, we introduce collaborative memory, a paradigm that connects isolated semantics to enable the sharing of relational insights. Yet, naively utilizing collaborative memory causes severe context overload and introduces noise to downstream LLMs, alongside prohibitive computational costs. To resolve this, we propose MemRec, a framework that architecturally decouples memory management from reasoning. MemRec introduces a dedicated, lightweight language model LM_Mem to efficiently manage and synthesize a dynamic collaborative memory graph in the background. It provides only distilled, high-signal contexts to a downstream, heavyweight large language model (LLM_Rec) for the final recommendation. Extensive experiments on four benchmarks demonstrate that MemRec achieves state-of-the-art performance. Code: https://github.com/rutgerswiselab/memrecHomepage: https://memrec.weixinchen.com

Today's open vocabulary scene graph generation (OVSGG) extends traditional SGG by recognizing novel objects and relationships beyond predefined categories, leveraging the knowledge from pre-trained large-scale models. Most existing methods adopt a two-stage pipeline: weakly supervised pre-training with image captions and supervised fine-tuning (SFT) on fully annotated scene graphs. Nonetheless, they omit explicit modeling of interacting objects and treat all objects equally, resulting in mismatched relation pairs. To this end, we propose an interaction-aware OVSGG framework INOVA. During pre-training, INOVA employs an interaction-aware target generation strategy to distinguish interacting objects from non-interacting ones. In SFT, INOVA devises an interaction-guided query selection tactic to prioritize interacting objects during bipartite graph matching. Besides, INOVA is equipped with an interaction-consistent knowledge distillation to enhance the robustness by pushing interacting object pairs away from the background. Extensive experiments on two benchmarks (VG and GQA) show that INOVA achieves state-of-the-art performance, demonstrating the potential of interaction-aware mechanisms for real-world applications.

Collaborative filtering (CF) is a fundamental technique in recommender systems, yet utilizing the vast amount of unlabeled data effectively poses a significant challenge. Recent research endeavors have concentrated on extracting subsets of this data that approximate negative samples. Regrettably, the remaining data are overlooked, failing to fully integrate this valuable information into the construction of user preferences. To bridge this gap, we propose a novel positive-neutral-negative learning paradigm (PNNP). PNNP introduces a neutral class that includes complex items that are difficult to categorize directly as positive or negative. By training a model based on this triple-wise partial ranking, PNNP provides a promising avenue for learning intricate user preferences. Through theoretical analysis, we connect PNNP to the one-way partial AUC (OPAUC) to validate its effectiveness. Implementing the PNN paradigm is, however, technically challenging because: (1) Modeling Neutral Samples: Users’ attitudes towards items classified as neutral can be complex and uncertain, requiring advanced modeling techniques. (2) Classifying Unlabeled Data: Without supervised signals, distinguishing between neutral and negative samples within unlabeled data is particularly challenging. (3) Lack of Suitable Loss Functions: There is no existing loss function that effectively manages set-level triple-wise ranking relationships. To address these challenges, we propose an innovative method to model neutral samples through the lens of uncertainty. Instead of representing neutral samples as fixed points in a high-dimensional space, we use Elliptical Gaussian Distributions to encapsulate their inherent uncertainty effectively. We then introduce a semi-supervised learning method combined with a user-aware attention model for enhanced knowledge acquisition and classification refinement. Furthermore, a novel loss function with a two-step centroid ranking approach is developed to handle set-level rankings. Extensive experiments on four real-world datasets demonstrate that, when integrated with PNNP, a broad range of representative CF models can consistently and significantly enhance their performance. Even a simple matrix factorization model, when combined with PNNP, can achieve performance comparable to sophisticated graph neural networks. Our code is publicly available at https://github.com/Asa9aoTK/PNN-RecBole.

Intuitively, an ideal collaborative filtering (CF) model should learn from users' full rankings over all items to make optimal top-K recommendations. Due to the absence of such full rankings in practice, most CF models rely on pairwise loss functions to approximate full rankings, resulting in an immense performance gap. In this paper, we provide a novel analysis using the multiple ordinal classification concept to reveal the inevitable gap between a pairwise approximation and the ideal case. However, bridging the gap in practice encounters two formidable challenges: (1) none of the real-world datasets contains full ranking information; (2) there does not exist a loss function that is capable of consuming ranking information. To overcome these challenges, we propose a pseudo-ranking paradigm (PRP) that addresses the lack of ranking information by introducing pseudo-rankings supervised by an original noise injection mechanism. Additionally, we put forward a new ranking loss function designed to handle ranking information effectively. To ensure our method's robustness against potential inaccuracies in pseudo-rankings, we equip the ranking loss function with a gradient-based confidence mechanism to detect and mitigate abnormal gradients. Extensive experiments on four real-world datasets demonstrate that PRP significantly outperforms state-of-the-art methods.

The effectiveness of a blended approach integrating face-to-face and eHealth interventions for promoting positive lifestyle behaviours in older adults has not been systematically tested. This study aimed to assess the feasibility and effectiveness of such interventions in improving health behaviours and outcomes among older adults in Hong Kong. A 10-week, single-blind, randomized controlled trial recruited 132 eligible older adults. Participants were assigned to three groups: (1) a blended intervention group: two sessions per week for ten weeks with one for physical activity and one for diet (fruit and vegetable intake; meat, fish, egg and alternatives intake) and two web-based sessions; (2) a face-to-face intervention group: same content and intensity like the blended group but as face-to-face sessions; and (3) a control group receiving biweekly telephone calls. Data on lifestyle behaviours and health outcomes (physical fitness, depression, loneliness, health-related quality of life) were collected at baseline (T1), 10 weeks post-test (T2), and a 3-month follow-up (T3). All data were analyzed using IBM SPSS 29.0. Descriptive statistics were used to describe the feasibility. The generalized linear mixed models were used to evaluate the effects of the intervention. The study demonstrated high feasibility with > 90

Generative retrieval-based recommendation has emerged as a promising paradigm aiming at directly generating the identifiers of the target candidates. However, in large-scale recommendation systems, this approach becomes increasingly cumbersome due to the redundancy and sheer scale of the token space. To overcome these limitations, recent research has explored the use of semantic tokens as an alternative to ID tokens, which typically leveraged reconstruction-based strategies, like RQ-VAE, to quantize content embeddings and significantly reduce the embedding size. However, reconstructive quantization aims for the precise reconstruction of each item embedding independently, which conflicts with the goal of generative retrieval tasks focusing more on differentiating among items. Moreover, multi-modal side information of items, such as descriptive text and images, geographical knowledge in location-based recommendation services, has been shown to be effective in improving recommendations by providing richer contexts for interactions. Nevertheless, effectively integrating such complementary knowledge into existing generative recommendation frameworks remains challenging. To overcome these challenges, we propose a novel unsupervised deep quantization exclusively based on contrastive learning, named SimCIT (a Simple Contrastive Item Tokenization framework). Specifically, different from existing reconstruction-based strategies, SimCIT propose to use a learnable residual quantization module to align with the signals from different modalities of the items, which combines multi-modal knowledge alignment and semantic tokenization in a mutually beneficial contrastive learning framework. Extensive experiments across public datasets and a large-scale industrial dataset from various domains demonstrate SimCIT's effectiveness in LLM-based generative recommendation.

Cross-domain recommendation (CDR) methods predominantly leverage overlapping users to transfer knowledge from a source domain to a target domain. However, through empirical studies, we uncover a critical bias inherent in these approaches: while overlapping users experience significant enhancements in recommendation quality, non-overlapping users benefit minimally and even face performance degradation. This unfairness may erode user trust, and, consequently, negatively impact business engagement and revenue. To address this issue, we propose a novel solution that generates virtual source-domain users for non-overlapping target-domain users. Our method utilizes a dual attention mechanism to discern similarities between overlapping and non-overlapping users, thereby synthesizing realistic virtual user embeddings. We further introduce a limiter component that ensures the generated virtual users align with real-data distributions while preserving each user's unique characteristics. Notably, our method is model-agnostic and can be seamlessly integrated into any CDR model. Comprehensive experiments conducted on three public datasets with five CDR baselines demonstrate that our method effectively mitigates the CDR non-overlapping user bias, without loss of overall accuracy. Our code is publicly available at https://github.com/WeixinChen98/VUG.

Emotional Support Conversation (ESC) systems aim to alleviate users' emotional difficulties and provide long-term, systematic support for emotional well-being. However, most large language model (LLM)-based ESC systems rely on predefined strategies, which limits their effectiveness in complex, real-life scenarios. To enable flexible responses to diverse emotional problem scenarios, this paper introduces a novel end-to-end framework (RLFF-ESC) that directly learns enduring emotionally supportive response skills using reinforcement learning. For sustained emotional support, we first employ an LLM-based multi-agent mechanism to simulate future dialogue trajectories and collect future-oriented rewards. We then train a future-oriented reward model, which is subsequently used to train the emotional support policy model. Additionally, we incorporate an explicit reasoning process during response generation to further enhance the quality, relevance, and contextual appropriateness of the system's responses. We evaluate the backbone policy model on Qwen2.5-7B-Instruct-1M and LLaMA3.1-8B-Instruct models, testing the proposed RLFF-ESC framework across two public ESC datasets. Experimental results demonstrate that RLFF-ESC consistently outperforms existing baselines in terms of goal completion and response quality.