Sequential recommendation (SR) has received much attention and made promising progress in the past few years due to its high alignment with real recommendation scenarios. It models users' preferences and behavior patterns from their historical behavior sequences and provides personalized recommendations. However, the widespread problem of data sparsity limits the performance of sequential recommendation models. To tackle this, data augmentation (DA) provides a feasible solution by improving the quantity, quality, or diversity of the training samples without the need for additional data collection. In this survey, we present a systematic and timely review of research efforts on data augmentation for sequential recommendation. We start by providing a clear formulation of the problem and task. Then, we develop a unified taxonomy that categorizes existing augmentation methodologies regarding their augmentation objects and principles. Next, we conduct a comparative discussion on the advantages and disadvantages of different categories, supplemented with quantitative performance evaluations, time-complexity analyses, and visual case studies of representative methods, aiming to provide actionable guidance for the selection and development of augmentation methods in real-world scenarios. Finally, we present the future research directions and summarize this survey.

Test-time reinforcement learning generates multiple candidate answers via repeated rollouts and performs online updates using pseudo-labels constructed by majority voting. To reduce overhead and improve exploration, prior work introduces tree-structured rollouts, which share reasoning prefixes and branch at key nodes to improve sampling efficiency. However, this paradigm still faces two challenges: (1) high-entropy branching can trigger rollout collapse, where the branching budget concentrates on a few trajectories with consecutive high-entropy segments, rapidly reducing the number of effective branches; (2) early pseudo-labels are noisy and biased, which can induce self-reinforcing overfitting, causing the policy to sharpen prematurely and suppress exploration. To address these issues, we propose Entropy–Confidence Hybrid Group Relative Policy Optimization (ECHO). During rollout, ECHO jointly leverages local entropy and group-level confidence to adaptively control branch width, and further introduces online confidence-based pruning to terminate persistently low-confidence branches, avoiding high-entropy traps and mitigating collapse. During policy updates, ECHO employs confidence-adaptive clipping and an entropy–confidence hybrid advantage shaping approach to enhance training robustness and mitigate early-stage bias. Experiments demonstrate that ECHO achieves consistent gains on multiple mathematical and visual reasoning benchmarks, and generalizes more effectively under a limited rollout budget.

Background: Early and accurate detection of malignant biliary strictures (MBSs) is critical for improving the clinical outcomes of patients with cholangiocarcinoma. Although digital single-operator cholangioscopy (dSOC) provides direct visualization, its diagnostic consistency is limited due to subjective interpretation. Methods: We constructed a convolutional neural network-based synergistic diagnostic system using dSOC images or videos along with clinical data from three tertiary centers in China. The model was trained using a primary development dataset and subsequently validated using independent internal and external cohorts. The model’s performance was compared with that of expert endoscopists and standard endoscopic retrograde cholangiography (ERCP)-based sampling. Findings: 497 patients with available cholangioscopy data from three academic centers were included to construct the Multimodal Cholangioscopy Algorithm for MBS Diagnostics (MCAM-Dx) in the study. The visual backbone (ResNet-50) achieved an area under the receiver operating characteristic curve (AUROC) of 0.88 for single-modality image analysis. The multimodal MCAM-Dx significantly outperformed single-modality models, achieving AUROCs of 0.86–0.98 and accuracies of 0.77–0.89 across validation cohorts. In the reader study, MCAM-Dx outperformed novice endoscopists and its performance was comparable to that of experts. In an independent video dataset of 76 patients, MCAM-Dx exhibited superior accuracy when compared with standard ERCP-based sampling, thereby supporting its integration into the cholangioscopy system for real-time assessment. Interpretation: MCAM-Dx consistently outperformed standard ERCP‑based sampling for the evaluation of MBSs, and its performance was comparable to that of expert endoscopists. As a translational extension of the primary diagnostic framework, we further integrated the trained system into a real‑time assistive prototype to determine its feasibility for clinical application.

Data augmentation is an effective technique for tackling data sparsity in sequential recommendation (SR). Existing methods generate new data during the model training to improve the performance. However, deploying them on a backbone model requires retraining, architecture modification, or introducing additional modules and learnable parameters. These processes are time-consuming and costly for well-trained models, especially when the model and data scales become large. In this work, we explore the test-time augmentation (TTA) for SR, which augments the input sequences during the inference phase and then fuses the model's predictions to improve final accuracy. It avoids the significant overhead associated with training-time augmentation. We first experimentally examine the potential of existing augmentation operators for TTA and find that the Substitute and Mask consistently achieve better performance. Further analysis reveals that these two operators retain the original sequential pattern while adding appropriate perturbations. Moreover, the random selection of augmentation positions creates suitable augmented samples from both semantic and temporal perspectives. Meanwhile, we find that the fixed operation ratio limits the diversity of augmented data, and the TTA may impair the model's performance on long sequences. In addition, the two operators still face time-consuming similarity-based item selection or interference from mask tokens. Based on the analysis and limitations, we present TNoise and TMask. The former injects uniform noise into the representation, avoiding the computational overhead of item selection. The latter blocks mask tokens from participating in model calculations (TMask-B) or directly removes interactions that should have been replaced with mask tokens (TMask-R). Further, we sample the augmentation ratio from a uniform distribution to improve the data diversity. For short sequences, we introduce a sequence smoothing and lengthening method based on inter-item interpolation. For long sequences, we set a threshold to avoid the negative effects of TTA. Comprehensive experiments demonstrate the effectiveness, efficiency, and generalizability of our method.

Natural language user preferences provide an interpretable interface for LLM personalization. However, universal preference summaries often contain information irrelevant to a particular downstream task. Directly supplying the full preference summary therefore wastes context capacity and introduces cross-task distraction, while manually designing task-specific preference views is difficult to scale. In this work, we study task-specific preference adaptation: given a universal user preference summary and a downstream task, derive a task-conditioned representation that preserves sufficient decision-relevant evidence while removing redundant context. To this end, we propose AlignXada, a training-free meta-learning framework that induces reusable textual refinement policies for adapting universal preference summaries to task-specific ones. The refinement policy is iteratively optimized by a meta learner through verbal reinforcement learning. Across 13 tasks and three downstream models (39 task–model cells), AlignXada achieves an average gain of 3.82 points, improving 33 cells while retaining only 22.8% of the original profile tokens and outperforming RAG in 36 cells. An extended faithfulness analysis further shows that the refined profiles remain largely grounded in the source preferences while preserving task-relevant personalization signals, suggesting that profile-side adaptation serves as a practical complement to universal memory construction for lifelong personalized agents.

Personalized image generation is crucial for improving the user experience, as it renders reference images into preferred ones according to user visual preferences. Although effective, existing methods face two main issues. First, existing methods treat all items in the user historical sequence equally when extracting user preferences, overlooking the varying semantic similarities between historical items and the reference item. Disproportionately high weights for low-similarity items distort users' visual preferences for the reference item. Second, existing methods heavily rely on consistency between generated and reference images to optimize the generation, which leads to underfitting user preferences and hinders personalization. To address these issues, we propose Retrieval Augment Personalized Image GenerAtion guided by Recommendation (RAGAR). Our approach uses a retrieval mechanism to assign different weights to historical items according to their similarities to the reference item, thereby extracting more refined users' visual preferences for the reference item. Then we introduce a novel rank task based on the multi-modal ranking model to optimize the personalization of the generated images instead of forcing depend on consistency. Extensive experiments and human evaluations on three real-world datasets demonstrate that RAGAR achieves significant improvements in both personalization and semantic metrics compared to five baselines.

Medication recommender is to suggest appropriate medication combinations based on a patient’s health history, e.g., diagnoses and procedures. Existing works represent different diagnoses/procedures well separated by one-hot encodings. However, they ignore the latent hierarchical structures of these medical terms, undermining the generalization performance of the model. For example, “Respiratory Diseases”, “Chronic Respiratory Diseases” and “Chronic Bronchiti” have a hierarchical relationship, progressing from general to specific. To address this issue, we propose a novel hierarchical encoder named HIER to hierarchically represent diagnoses and procedures, which is based on standard medical codes and compatible with any existing methods. Specifically, the proposed method learns relation embedding with a self-supervised objective for incorporating the neighbor hierarchical structure. Additionally, we develop the position encoding to explicitly introduce global hierarchical position. Extensive experiments demonstrate significant and consistent improvements in recommendation accuracy across four baselines and two real-world clinical datasets. The code is available at https://github.com/yuliang-liang/HEIR .

Recently, sign-aware graph recommendations have drawn attention as they learn users’ negative preferences in addition to positive ones. Nevertheless, due to adopting two independent encoders for positive and negative interactions, existing approaches fail to learn the users’ comprehensive negative preferences and holistic collaborative signals from high-order heterogeneous interactions formed by multiple links with different signs. To compensate for this drawback, we devise a novel unified modeling approach to capture complete collaborative information and comprehensive user preferences. In this paper, we first explore the relationship between negative preferences and find that propagating both positive and negative high-order preferences along positive edges is feasible. Based on the observation, a Light Signed Graph Convolution Network for Recommendation (LSGRec) is proposed to comprehend user preferences within signed user-item interaction graphs. Then, recommendation results are generated based on positive preferences and optimized with negative ones. Finally, representations of users and items are trained through different auxiliary tasks. Extensive experiments on three real-world datasets demonstrate that our method outperforms existing baselines regarding performance and computational efficiency.

We study the problem of personalization in large language models (LLMs). Prior work predominantly represents user preferences as implicit, model-specific vectors or parameters, yielding opaque “black-box” profiles that are difficult to interpret and transfer across models and tasks. In contrast, we advocate natural language as a universal, model- and task-agnostic interface for preference representation. The formulation leads to interpretable and reusable preference descriptions, while naturally supporting continual evolution as new interactions are observed. To learn such representations, we introduce a two-stage training framework that combines supervised fine-tuning on high-quality synthesized data with reinforcement learning to optimize long-term utility and cross-task transferability. Based on this framework, we develop AlignXplore+, a universal preference reasoning model that generates textual preference summaries. Experiments on nine benchmarks show that our 8B model achieves state-of-the-art performanc – outperforming substantially larger open-source models – while exhibiting strong transferability across tasks, model families, and interaction formats.

Network measurement sketches enable efficient traffic monitoring but require careful parameter configuration to balance accuracy and memory efficiency. We present RA-Sketch, a unified framework for generating memory-optimal sketch configurations that satisfy user-defined error constraints across diverse network measurement tasks. Unlike existing approaches that rely on computationally intensive experimental testing, RA-Sketch introduces: 1) Poisson-distributed collision modeling to construct error predictors for both frequency-independent tasks (membership query, heavy-hitter detection, and super-spreader detection) and frequency-dependent tasks (flow size distribution, frequency estimation, and cardinality estimation), eliminating the need for empirical validation; 2) A hierarchical search strategy combining power-of-two scaling and binary search, reducing iterations through optimized parameter initialization. RA-Sketch supports 10+ sketch architectures including Bloom Filter, Elastic Sketch, HeavyKeeper, MEC Sketch, MRAC, CM Sketch, CO Sketch, gSkt, rSkt1 among others. Evaluations on real-world network traces demonstrate: 1) up to 6-7 orders-of-magnitude faster configuration than benchmark-based methods; 2) Prediction errors are within 10% for heavy-hitter detection and super-spreader detection in most evaluated settings, while prediction errors for membership query, flow size distribution, frequency estimation, and cardinality estimation are close to zero; 3) Memory utilization approaches theoretical minima. The framework's generality and efficiency enable real-time reconfiguration of sketches under dynamic network conditions.

Hard negative sampling improves recommendation performance by accelerating convergence and sharpening the decision boundary. However, most existing methods rely on heuristic strategies, selecting negatives from a fixed candidate pool. Lacking semantic awareness, these methods often misclassify items that align with users' semantic interests as negatives, resulting in False Hard Negative Samples (FHNS). Such FHNS inject noisy supervision and hinder the model's optimal performance. To address this challenge, we propose HNLMRec, a generative semantic negative sampling framework. Leveraging the semantic reasoning capabilities of Large Language Models (LLMs), HNLMRec directly generates negative samples that are behaviorally distinct yet semantically relevant with respect to user preferences. Furthermore, we integrate collaborative filtering signals into the LLM via supervised fine-tuning, guiding the model to synthesize more reliable and informative hard negatives. Extensive experiments on multiple real-world datasets demonstrate that HNLMRec significantly outperforms traditional methods and LLM-enhanced baselines, while effectively mitigating popularity bias and data sparsity, thereby improving generalization.

In modern data centers and enterprise networks, software switches have become critical components for achieving flexible and efficient network management. Due to resource constraints in software switches, sketches have emerged as a promising approach for network traffic measurement. However, their accuracy is often impacted by hash collisions. Existing hash functions treat all collisions equally, failing to account for the differing impacts of collisions involving elephant flows versus mouse flows. We propose FSA-Hash, a novel flow-size-aware hashing scheme that separates elephant flows from each other and from mouse flows, minimizing the most detrimental collisions. FSA-Hash is designed based on two insights: separating elephant flows from mouse flows avoids overestimating mouse flows, while separating elephant flows from each other enables accurate heavy-hitter detection. We implement FSA-Hash using machine learning models trained on network traffic data (LFSA-Hash), and also design a lightweight online variant (OLFSA-Hash) that learns the hash model solely from sketch queries on the software switch, obviating traffic collection overheads. Evaluations across four sketches and two tasks demonstrate FSA-Hash’s superior accuracy over standard hash functions. Moreover, OLFSA-Hash closely matches LFSA-Hash’s performance, making it an attractive option for adaptively refining the hash model without monitoring traffic.

Personalized text generation aims to infer users’ writing style preferences from their historical texts and generate outputs that faithfully reflect these stylistic characteristics. Existing solutions primarily adopt two paradigms: retrieval-augmented generation (RAG) and parameter-efficient fine-tuning (PEFT). While these approaches have advanced the field, they suffer from two critical limitations: (1) the entanglement of content semantics and stylistic patterns in historical texts impedes accurate modeling of user-specific writing preferences; and (2) scalability challenges arising from both RAG’s inference latency by retrieval operations and PEFT’s parameter storage requirements for per user model. To overcome these limitations, we propose StyleVector, a training-free framework that disentangles and represents personalized writing style as a vector in LLM’s activation space, enabling style-steered generation during inference without requiring costly retrieval or parameter storage. Comprehensive experiments demonstrate that our framework achieves a significant 8% relative improvement in personalized generation while reducing storage requirements by 1700 × over PEFT method.

By generating new yet effective data, data augmentation has become a promising method to mitigate the data sparsity problem in sequential recommendation. Existing works focus on augmenting the original data but rarely explore the issue of imbalanced relevance and diversity for augmented data, leading to semantic drift problems or limited performance improvements. In this paper, we propose a novel Balanced data Augmentation Plugin for Sequential Recommendation (BASRec) to generate data that balance relevance and diversity. BASRec consists of two modules: Single-sequence Augmentation and Cross-sequence Augmentation. The former leverages the randomness of the heuristic operators to generate diverse sequences for a single user, after which the diverse and the original sequences are fused at the representation level to obtain relevance. Further, we devise a reweighting strategy to enable the model to learn the preferences based on the two properties adaptively. The Cross-sequence Augmentation performs nonlinear mixing between different sequence representations from two directions. It produces virtual sequence representations that are diverse enough but retain the vital semantics of the original sequences. These two modules enhance the model to discover fine-grained preferences knowledge from single-user and cross-user perspectives. Extensive experiments verify the effectiveness of BASRec. The average improvement is up to 72.0% on GRU4Rec, 33.8% on SASRec, and 68.5% on FMLP-Rec. We demonstrate that BASRec generates data with a better balance between relevance and diversity than existing methods.

Network measurement sketches enable efficient traffic monitoring but require careful parameter configuration to balance accuracy and memory efficiency. We present RA-Sketch, a framework for generating memory-optimal sketch configurations that satisfy user-defined error constraints across diverse network measurement tasks. Unlike existing approaches that rely on computationally intensive experimental testing, RA-Sketch introduces: 1) Poisson-distributed collision modeling to construct error predictors for both frequency-independent tasks (membership query, heavy-hitter detection) and frequency-dependent tasks (frequency/cardinality estimation), eliminating the need for empirical validation; 2) A hierarchical search strategy combining power-of-two scaling and binary search, reducing iterations through optimized parameter initialization. RA-Sketch supports 10+ sketch architectures including Bloom Filter, Elastic Sketch, HeavyGuardian, HeavyKeeper, CM/CO Sketch, gSkt, rSkt1 and so on. Evaluations on real-world network traces demonstrate: 1) 6–7 orders of magnitude faster configuration than benchmark-based methods; 2) Prediction errors ≤10% for heavy-hitter detection, while prediction errors for membership query, and frequency/cardinality estimation are close to zero; 3) Memory utilization approaches theoretical minima. The framework’s generality and efficiency enable real-time reconfiguration of sketches under dynamic network conditions.

Network traffic measurement is critical for effective network management. Sketch has been proven to be a promising network traffic measurement solution. Considering the skewed distribution of network traffic, where low-frequency mouse flows dominate and high-frequency elephant flows are fewer, recent sketch-based solutions employ hierarchical designs to enhance memory efficiency and accuracy. However, these solutions inevitably introduce additional challenges, including increased memory access overhead, severe hash collisions between elephant and mouse flows, and limited adaptability to dynamic network environments. In this paper, we propose LA-Sketch, an adaptive level-aware data structure. First, LA-Sketch employs a level-aware classifier to intelligently map each flow to its corresponding level, thereby reducing memory access overhead caused by hierarchical designs and mitigating hash collisions between elephant and mouse flows. Second, we introduce an adaptive counter configuration method that dynamically adjusts the number of counters at each level according to diverse network traffic distributions, which theoretically minimizes overall hash collisions. Finally, to adapt to the continuously changing network traffic characteristics, we propose an adaptive online training method that enables LA-Sketch's classifier to maintain high performance using only sketch query values for training, avoiding the significant overhead of massive traffic data collection. Extensive evaluations on two real-world network traces across five measurement tasks demonstrate that LA-Sketch outperforms state-of-the-art hierarchical sketches.

Graph Neural Networks (GNNs)-based recommendation algorithms typically assume that training and testing data are drawn from independent and identically distributed (IID) spaces. However, this assumption often fails in the presence of out-of-distribution (OOD) data, resulting in significant performance degradation. In this study, we construct a Structural Causal Model (SCM) to analyze interaction data, revealing that environmental confounders (e.g., the COVID-19 pandemic) lead to unstable correlations in GNN-based models, thus impairing their generalization to OOD data. To address this issue, we propose a novel approach, graph representation learning via causal diffusion (CausalDiffRec) for OOD recommendation. This method enhances the model’s generalization on OOD data by eliminating environmental confounding factors and learning invariant graph representations. Specifically, we use backdoor adjustment and variational inference to infer the real environmental distribution, thereby eliminating the impact of environmental confounders. This inferred distribution is then used as prior knowledge to guide the representation learning in the reverse phase of the diffusion process to learn the invariant representa- tion. In addition, we provide a theoretical derivation that proves optimizing the objective function of CausalDiffRec can encourage the model to learn environment-invariant graph representations, thereby achieving excellent generalization performance in recom- mendations under distribution shifts. Our extensive experiments validate the effectiveness of CausalDiffRec in improving the generalization of OOD data, and the average improvement is up to 10.69% on Food, 18.83% on KuaiRec, 22.41% on Yelp2018, and 11.65% on Douban datasets.

Prompt tuning is a promising method to fine-tune a pre-trained language model without retraining its large-scale parameters. Instead, it attaches a soft prompt to the input text, whereby downstream tasks can be well adapted by merely learning the embeddings of prompt tokens. Nevertheless, existing methods still suffer from two challenges: (i) they are hard to balance accuracy and efficiency. A longer (shorter) soft prompt generally leads to a better (worse) accuracy but at the cost of more (less) training time. (ii) The performance may not be consistent when adapting to different downstream tasks. We attribute it to the same embedding space but responsible for different requirements of downstream tasks. To address these issues, we propose an Efficient Prompt Tuning method (EPT) by multi-space projection and prompt fusion. Specifically, it decomposes a given soft prompt into a shorter prompt and two low-rank matrices, significantly reducing the training time. Accuracy is also enhanced by leveraging low-rank matrices and the short prompt as additional knowledge sources to enrich the semantics of the original short prompt. In addition, we project the soft prompt into multiple subspaces to improve the performance consistency, and then adaptively learn the combination weights of different spaces through a gating network. Experiments on 13 natural language processing downstream tasks show that our method significantly and consistently outperforms 11 comparison methods with the relative percentage of improvements up to 12.9%, and training time decreased by 14%.

Multimodal recommendation aims to model user and item representations comprehensively with the involvement of multimedia content for effective recommendations. Existing research has shown that it is beneficial for recommendation performance to combine (user- and item-) ID embeddings with multimodal salient features, indicating the value of IDs. However, there is a lack of a thorough analysis of the ID embeddings in terms of semantics in the literature. In this paper, we revisit the value of ID embeddings for multimodal recommendation and conduct a thorough study regarding its semantics, which we recognize as subtle features of content and structure. Based on our findings, we propose a novel recommendation model by incorporating ID embeddings to enhance the salient features of both content and structure. Extensive experiments on three real-world datasets (Baby, Sports, and Clothing) demonstrate the superiority of our method over state-of-the-art multimodal recommendation methods and the effectiveness of fine-grained ID embeddings.

Recommendation unlearning is an emerging task to erase the influences of user-specified data from a trained recommendation model. Most existing research follows the paradigm of partitioning the original dataset into multi-fold and then retraining corresponding sub-models while those influences are totally removed. Despite the effectiveness, two key problems remain unexplored: i) Existing work becomes inefficient and computationally expensive to retrain all sub-models, especially when facing large amounts of unlearning data. ii) User preferences are dynamically changing. If users express negative opinions on some interacted items they used to prefer, how can we adaptively erase the outdated preferences behind such transformation from the trained model? Although these unlearning data contain outdated information, there is still a lot of helpful knowledge worth preserving. Existing methods ignore this preservation during unlearning and may remove all the knowledge in the interactions, compromising the final performance. In light of these limitations, we propose a novel unlearning framework called GFEraser, which transforms the unlearning into an efficient guided filtering process to avoid time-consuming retraining and retain beneficial knowledge. Specifically, we develop an intra-user negative sampling strategy to learn the outdated preferences that need to be erased. Under the guidance of differential maximization agreement and attention-based fusion module, the original representations are adaptively filtered and aggregated based on the learned preferences. Besides, we leverage contrastive learning to preserve the invariant user preferences, maintaining the final performance. Finally, we devise a new metric called Ranking Decrease Rate to evaluate the unlearning effect. Experimental results demonstrate that GFEraser can maintain reliable recommendation performance while achieving efficient outdated preferences unlearning, up to 37x acceleration.