Recommender systems have become an essential component of many online platforms, providing personalized recommendations to users. A crucial aspect is embedding techniques that convert the high-dimensional discrete features, such as user and item IDs, into low-dimensional continuous vectors, which can enhance the recommendation performance. Embedding techniques have revolutionized the capture of complex entity relationships, generating significant research interest. This survey presents a comprehensive analysis of recent advances in recommender system embedding techniques. We examine centralized embedding approaches across matrix, sequential, and graph structures. In matrix-based scenarios, collaborative filtering generates embeddings that effectively model user-item preferences, particularly in sparse data environments. For sequential data, we explore various approaches including recurrent neural networks and self-supervised methods such as contrastive and generative learning. In graph-structured contexts, we analyze techniques like node2vec that leverage network relationships, along with applicable self-supervised methods. Our survey addresses critical scalability challenges in embedding methods and explores innovative directions in recommender systems. We introduce emerging approaches, including AutoML, hashing techniques, and quantization methods, to enhance performance while reducing computational complexity. Additionally, we examine the promising role of Large Language Models (LLMs) in embedding enhancement. Through detailed discussion of various architectures and methodologies, this survey aims to provide a thorough overview of state-of-the-art embedding techniques in recommender systems, while highlighting key challenges and future research directions. To facilitate development, evaluation, and comparison of embedding-based recommender systems, we provide an open source repository ( https://github.com/Applied-Machine-Learning-Lab/Embedding-in-Recommender-Systems ).

Graph Neural Networks (GNNs) learn from graph-structured data by passing messages between neighboring nodes along edges on certain topological layouts. While layouts can be essential to GNNs' performance, extant methods generally consider obtaining layouts from limited perspectives. In this paper, we introduce Distributional Edge Layouts (DELs), a first-of-its-kind method to sample a collection of topological layouts from a Boltzmann distribution under physical energies. By integrating DELs into GNNs, a wide landscape of feasible graph layouts can be captured from a holistic perspective, overcoming the intrinsic drawbacks in existing GNN designs.In practice, DELs can complement various GNN architectures with high versatility. Our theoretical analysis proves that GNNs equipped with DELs maintain at least the same expressive as their original counterparts, with empirical potential offering extra expressivity. Extensive experiments demonstrate that DELs consistently and substantially improve the performance of a wide range of GNN baselines across multiple datasets, achieving state-of-the-art results. This improvement suggests that DELs capture important distributional information previously overlooked by traditional GNN approaches. DEL is open-sourced at https://github.com/LOGO-CUHKSZ/DEL.

Graph Neural Networks operate through bottom-up message-passing, fundamentally differing from human visual perception, which intuitively captures global structures first. We investigate the underappreciated potential of vision models for graph understanding, finding they achieve performance comparable to GNNs on established benchmarks while exhibiting distinctly different learning patterns. These divergent behaviors, combined with limitations of existing benchmarks that conflate domain features with topological understanding, motivate our introduction of GraphAbstract. This benchmark evaluates models' ability to perceive global graph properties as humans do: recognizing organizational archetypes, detecting symmetry, sensing connectivity strength, and identifying critical elements. Our results reveal that vision models significantly outperform GNNs on tasks requiring holistic structural understanding and maintain generalizability across varying graph scales, while GNNs struggle with global pattern abstraction and degrade with increasing graph size. This work demonstrates that vision models possess remarkable yet underutilized capabilities for graph structural understanding, particularly for problems requiring global topological awareness and scale-invariant reasoning. These findings open new avenues to leverage this underappreciated potential for developing more effective graph foundation models for tasks dominated by holistic pattern recognition.

Existing scheduling methods for redundant executors in mimic defense systems often lack dynamism and typically only consider second-order heterogeneity. This limitation makes systems vulnerable to attacks that exploit common-mode vulnerabilities. Furthermore, these methods fail to differentiate the varying threat levels posed by common-mode vulnerabilities of different orders, consequently overestimating low-order risks while underestimating high-order ones. To address these issues, this paper proposes an efficient dynamic scheduling algorithm based on weighted hypergraph heterogeneity, named HARS(Hypergraph-based Amplified Risk Scheduling). For the first time, we accurately model the executor-vulnerability relationships in a Dynamic Heterogeneous Redundancy (DHR) system as a weighted hypergraph, where hyperedges intuitively represent common-mode vulnerabilities. By introducing a risk weight function, our model achieves differentiated and amplified measurement of high-order security risks. The scheduling decision is then transformed into a graph-theoretic optimization problem that can be approximately solved in polynomial time. We further develop an efficient heuristic greedy algorithm, fundamentally resolving the exponential complexity bottleneck associated with high-order heterogeneity computation. Simulation results demonstrate that the proposed HARS algorithm not only inherits the negative feedback mechanism but also achieves a superior trade-off between security and system performance through more precise and efficient high-order heterogeneity calculations.

ABSTRACT With the rise of social media and the spread of a large number of pictures on the Internet, protecting data privacy and verifying copyright has become hot research. A common method is to use digital watermarking. However, the existing blind watermarking methods only consider embedding the watermark in the image itself and ignore the fact that the attacker can remove the watermark through other shooting devices. Therefore, to solve this problem, we propose an image watermarking method based on few‐shot learning. We use an autoencoder to learn the embedding and extraction of watermarks. Then, we propose a framework named Simulated Candid Shooting Layer (SCSL). The SCSL simulates a variety of candid scenes as noise data using meta‐learning and enhances the robustness of image watermarking. Experiments show that the proposed method is superior to the state of the art in watermarking technologies in both robustness and invisibility of watermarking. Specifically, it achieved an improvement of over 8% in evaluation metrics against JPEG attacks. The proposed SCSL framework further enhanced these metrics by more than 5%.

Recently, there has been an emerging trend to integrate persistent homology (PH) into graph neural networks (GNNs) to enrich expressive power. However, naively plugging PH features into GNN layers always results in marginal improvement with low interpretability. In this paper, we investigate a novel mechanism for injecting global topological invariance into pooling layers using PH, motivated by the observation that filtration operation in PH naturally aligns graph pooling in a cut-off manner. In this fashion, message passing in the coarsened graph acts along persistent pooled topology, leading to improved performance. Experimentally, we apply our mechanism to a collection of graph pooling methods and observe consistent and substantial performance gain over several popular datasets, demonstrating its wide applicability and flexibility.