In recent years, video recognition models have witnessed the rapid development of Deep Neural Networks (DNNs). However, these models remain not robust to adversarial examples that are created by adding imperceptible perturbations to clean samples. Recent studies indicate that generating adversarial examples in the hard-label black-box setting is particularly challenging yet highly practical. Compared to image recognition models, there are few hard-label black-box adversarial example generation algorithms for video recognition models. To this end, we propose a hard-label black-box video adversarial example generation algorithm, referred to as Dynamic Black-box Algorithm (DBA). First, DBA uses the binary search algorithm to find the boundary video between two original videos; then, the sampling-based algorithm is used to estimate the gradient on the boundary video; finally, with a dynamic step size adjustment strategy, DBA moves the boundary video towards the direction of the estimated gradient to generate the adversarial video. Additionally, we designed another strategy to skip invalid samples generated during the adversarial example generation process. Experiments demonstrate that DBA attains a superior trade-off between the magnitude of perturbations and query efficiency. Specifically, DBA outperforms state-of-the-art algorithms, achieving an average reduction in Mean Squared Error (MSE) of over 50%.

Combinational equivalence checking (CEC) is essential for verifying the correctness of circuit designs. With the growing complexity of circuits, effective verification techniques have become increasingly critical. Recently, a conjunctive normal form (CNF)-based approach, converting circuits to CNF for Boolean satisfiability (SAT) solvers, has shown competitive performance compared to state-of-the-art hybrid SAT sweeping approaches. The capability of this CNF-based approach depends on effective CNF conversion. This work presents CirOPT, which is the first CNF conversion method using compiler optimization to equivalently simplify circuits. Extensive experiments are conducted on a broad range of real-world benchmarks, which are far more than the number of benchmarks typically used in empirical studies. The results reveal that, when paired with the state-of-the-art CNF SAT solver Kissat, CirOPT considerably outperforms existing approaches in CEC.

Graph embedding represents each graph in a low-dimensional space with similarity between graph pairs preserved. While the mainstream Graph Neural Networks (GNNs) achieve strong performance, they pose significant computational challenges and predominantly focus on discrete-attribute graphs. In this paper, we propose #WLS, a learning-free continuous-attribute graph embedding model that keeps a good trade-off between accuracy and efficiency by employing Locality-Sensitive Hashing (LSH) to preserve high-order node similarity. Experimental results on seven real-world datasets (405 to 41,127 graphs) show that #WLS achieves accuracy comparable to representative GNN methods in graph classification (e.g., 80.12% vs. 76.83% on OGBG_MOLHIV) while reducing runtime (up to 27,183 & times; speedups in our experiments) and maintaining a low memory footprint (under 300MB on PROTEINS_full and AIDS). It also outperforms existing LSH-based methods on most datasets. In graph retrieval, #WLS attains MAP scores competitive with GNN methods (e.g., 74.27% vs. 72.25% on PROTEINS_full) and outperforms existing LSH-based methods across the evaluated datasets.

Adaptive Traffic Signal Control (ATSC) is a pivotal research area within intelligent transportation systems, aiming to enhance transportation efficiency and alleviate traffic congestion at signalized intersections. While multi-agent deep reinforcement learning has been extensively applied to ATSC, existing approaches commonly frame it as a fully cooperative problem, presupposing that all agents are committed to pursuing a collective optimal solution. However, achieving such altruistic cooperation is often impractical. Furthermore, as the number of agents escalates, challenges such as the curse of dimensionality and non-stationarity arise, complicating the learning process. To address these issues, we propose a novel perspective by framing ATSC as a competitive-cooperative game trade-off scenario and design a multi-agent framework, termed Neighborhood Coordinated and Holistic Optimized Actor-Critic (NcHo-AC). Specifically, we introduce a novel traffic state representation, design a sophisticated feature extraction network, develop a robust training algorithm, and leverage mean field approximation to model population-level agent interactions. These designs foster neighborhood-level cooperation and communication, facilitate the learning of the desired Nash equilibrium, and mitigate the noise caused by agents’ exploratory behaviors, thereby alleviating non-stationarity and the curse of dimensionality, while enhancing scalability to large-scale traffic networks. Comprehensive experiments conducted on both synthetic and real-world datasets demonstrate that NcHo-AC significantly outperforms state-of-the-art baselines across four key metrics: average travel time, average queue length, delay, and throughput, along with improved convergence, robustness, and interpretability.

Set covering problem (SCP) is a fundamental combinatorial optimization problem in computer science, and it has been extensively applied in real-world, industrial scenarios. Given a universal set of items X and a collection of subsets Y ? 2 X , where each subset y j ∈ Y is a subset of X , and the union of all subsets in Y is equal to X , each subset y j ∈ Y is associated with a positive integer as its weight ω ( y j ) . The objective of SCP is to find a collection S ? Y with the minimum total weight to ensure that the union of all subsets in S equals X .

Attributed network embedding seeks to depict each network node via a compact, low-dimensional vector while effectively preserving the similarity between node pairs, which lays a strong foundation for a great many high-level network mining tasks. With the advent of the era of big data, the number of nodes and edges has reached billions in many real-world networks, which poses great computational and storage challenges to the existing methods. Although some algorithms have been developed to handle billion-scale networks, they often undergo accuracy degradation or tempo-spatial inefficiency owing to attribute information loss or substantial parameter learning. To this end, we propose a simple, time- and space-efficient billion-scale attributed network embedding algorithm called SketchBANE in this paper, which strikes an excellent balance between accuracy and efficiency by adopting sparse random projection with 1-bit quantization to sketch the iterative closed neighborhood and maintain the similarity among high-order nodes in a non-learning manner. The extensive experimental results indicate that our proposed SketchBANE algorithm competes favorably with the state-of-the-art approaches, while remarkably reducing runtime and space consumption. Also, the proposed SketchBANE algorithm exhibits good scalability and parallelization.

Real-world networks, particularly those in web and social media, are dynamic with evolving node attributes and structures, often involving billions of nodes and edges. Dynamic attributed network embedding is a powerful tool for capturing these changes, enabling data owners and problem owners to better understand interactions and trends for more effective engagement and decision-making. While some existing algorithms are capable of handling very large-scale dynamic attributed networks with billions of nodes and edges, they often suffer from accuracy loss or high computational overhead. In this paper, we propose a practical and sustainable framework of sketching very large-scale dynamic attributed networks called VLS(2)ketch, which incorporates incremental embedding updates alongside storage-efficient, binarized representation of both node attributes and topological variations. By the sparse random projection technique in an incremental update manner, VLS(2)ketch significantly reduces the energy-intensive computational workload while maintaining accuracy. Also, we introduce an information decay mechanism, which adapts to temporally varying topologies and node attributes. This mechanism ensures that outdated information gradually diminishes over time. Extensive experiments on real-world very large-scale datasets demonstrate that our proposed VLS(2)ketch method delivers comparable embedding quality against the state-of-the-art learning-based competitors with dramatically reduced runtime(1).

Highly configurable software systems are crucial in practice to satisfy the rising demand for software customization, and combinatorial interaction testing (CIT) is an important methodology for testing such systems. Constrained covering array generation (CCAG), as the core problem in CIT, is to construct a $t$-wise covering array (CA) of minimum size, where $t$ represents the testing strength. Extensive studies have demonstrated that high-strength CIT (e.g., 4-wise and 5-wise CIT) has stronger fault detection capability than low-strength CIT (i.e., 2-wise and 3-wise CIT), and there exist certain critical faults that can be disclosed through high-strength CIT. Although existing CCAG algorithm has exhibited effectiveness in solving the low-strength CCAG problem, they suffer the severe highstrength challenge when solving 4-wise and 5-wise CCAG, which urgently calls for effective solutions to solving 4-wise and 5 wise CCAG problems. To alleviate the high-strength challenge, we propose a novel and effective local search algorithm dubbed HSCA. Particularly, HSCA incorporates three new and powerful techniques, i.e., multi-round CA generation mechanism, dynamic priority assigning technique, and variable grouping strategy, to improve its performance. Extensive experiments on 35 real-world and synthetic instances demonstrate that HSCA can generate significantly smaller 4-wise and 5-wise CAs than existing state-of-the-art CCAG algorithms. More encouragingly, among all 35 instances, HSCA successfully builds 4-wise and 5-wise CAs for 35 and 29 instances, respectively, including 11 and 15 instances where existing CCAG algorithms fail. Our results indicate that HSCA can effectively mitigate the high-strength challenge.

This work focuses on effectively generating diverse solutions for satisfiability modulo theories (SMT) formulas, targeting the theories of bit-vectors, arrays, and uninterpreted functions, which is a critical task in software and hardware testing. Generating diverse SMT solutions helps uncover faults and detect safety violations during the verification and testing process, resulting in the SMT sampling problem, i.e., constructing a small number of solutions while achieving comprehensive coverage of the constraint space. While high coverage is crucial for exploring system behaviors, reducing the number of solutions is of great importance, as excessive solutions increase testing time and resource usage, undermining efficiency. In this work, we introduce PanSampler, a novel SMT sampler that achieves high coverage with a small number of solutions. It incorporates three novel techniques, i.e., diversity-aware SMT algorithm, abstract syntax tree (AST)-guided scoring function and post-sampling optimization technology, enhancing its practical performance. It iteratively samples solutions, evaluates candidates, and employs local search to refine solutions, ensuring high coverage with a small number of samples. Extensive experiments on practical benchmarks demonstrate that PanSampler exhibits a significantly stronger capability to reach high target coverage, while requiring fewer solutions than current samplers to achieve the same coverage level. Furthermore, our empirical evaluation on practical subjects, which are collected from real-world software systems, shows that PanSampler achieves higher fault detection capability and reduces the number of required test cases from 32.6% to 76.4% to reach the same fault detection effectiveness, leading to a substantial improvement in testing efficiency. PanSampler advances SMT sampling, reducing the cost of software testing and hardware verification.

How to detect DoS attacks in SDN precisely and intelligently is a big challenge. Most of SDN DoS attack detection methods artificially analyze DoS traffic characteristics, and thus they are unscalable. Additionally, they pay little attention to minimizing the impact of DoS attacks on other normal traffic of the victim switches, which might make packets from normal traffic hard to be served because DoS attack traffic jams the bandwidth between the victim switches and the controller. In this paper, we propose DADF, which is an intelligent framework deployed in the controller to detect DoS attacks in SDN. The framework has two important modules, i.e., the detection module and the diversion module. The detection module aims to detect DoS attacks precisely and intelligently by employing deep learning. The diversion module adopts reinforcement learning to migrate DoS attack traffic intelligently and minimize the impact of DoS attacks on other normal traffic of the victim switches. The extensive experimental evaluation shows that DADF can precisely detect DoS attacks and minimize the impact of DoS attacks on other normal traffic of the victim switches by migrating DoS attack traffic to neighbor switches intelligently.

Virtual Network Embedding (VNE) is a fundamental resource allocation challenge that is associated with hard and multifaceted constraints in network function virtualization (NFV). Existing works for VNE struggle to handle such complex constraints, leading to compromised system performance and stability. In this paper, we propose a \textbf{CON}straint-\textbf{A}ware \textbf{L}earning framework, named \textbf{CONAL}, for efficient constraint handling in VNE. Concretely, we formulate the VNE problem as a constrained Markov decision process with violation tolerance, enabling precise assessments of both solution quality and constraint violations. To achieve the persistent zero violation to guarantee solutions' feasibility, we propose a reachability-guided optimization with an adaptive reachability budget method. This method also stabilizes policy optimization by appropriately handling scenarios with no feasible solutions. Furthermore, we propose a constraint-aware graph representation method to efficiently learn cross-graph relations and constrained path connectivity in VNE. Finally, extensive experimental results demonstrate the superiority of our proposed method over state-of-the-art baselines. Our code is available at \href{https://github.com/GeminiLight/conal-vne}{https://github.com/GeminiLight/conal-vne}.

Two analogous tasks have emerged in multimodal dialogue research: multimodal dialogue response generation and multimodal task-oriented dialogue. Both tasks share the goal of response multi-round, interactive content based on the multimodal dialogue history, but the latter focuses on accomplishing specific objectives which can be viewed as the former fine-tuned. The fine-tuning strategy may cause catastrophic forgetting and overfitting on few well-annotated data. Despite considerable progress in both areas, many existing works rely on retrieval-based approaches and additional auxiliary knowledge bases. To address these issues, we propose a Hierarchical Fusion Framework (HFF) for multimodal dialogue response generation. HFF blends these two tasks to learn a generation model from a data-driven perspective by introducing multi-dataset learning scheme, achieving a balance between generalization and expertise. In this work, multi-dataset learning is cast as a multi-objective optimization problem due to potential conflicts between datasets, necessitating a trade-off based on data distribution during training. Hierarchical fusion is performed sequentially between modalities and datasets, which could efficiently establish clear cross-modal relationships and integrate knowledge from multi-dataset. Specifically, HFF aligns extracted unimodal features (image and text) before fusing them through cross-modal attention and integrates them into multimodal encoder-decoder for generating responses. By optimizing the fusion between corpora from multi-dataset as conflicting objectives to satisfy Pareto optimality, our approach effectively facilitates both multimodal task-oriented and task-unoriented dialogues. Experimental results demonstrate the effectiveness of HFF and its comparable performance with all baselines.

Given a network, it is well recognized that attributed network embedding represents each node of the network in a low-dimensional space, and, thus, brings considerable benefits for numerous graph mining tasks. In practice, a diverse set of graph tasks can be processed efficiently via the compact representation that preserves content and structure information. The majority of attributed network embedding approaches, especially, the graph neural network (GNN) algorithms, are substantially costly in either time or space due to the expensive learning process, while the randomized hashing technique, locality-sensitive hashing (LSH), which does not need learning, can speedup the embedding process at the expense of losing some accuracy. In this article, we propose the MPSketch model, which bridges the performance gap between the GNN framework and the LSH framework by adopting the LSH technique to pass messages and capture high-order proximity in a larger aggregated information pool from the neighborhood. The extensive experimental results confirm that in node classification and link prediction, the proposed MPSketch algorithm enjoys performance comparable to the state-of-the-art learning-based algorithms and outperforms the existing LSH algorithms, while running faster than the GNN algorithms by 3-4 orders of magnitude. More precisely, MPSketch runs 2121, 1167, and 1155 times faster than GraphSAGE, GraphZoom, and FATNet on average, respectively.

The Centralized Training and Decentralized Execution (CTDE) paradigm, where a centralized critic is allowed to access global information during the training phase while maintaining the learned policies executed with only local information in a decentralized way, has achieved great progress in recent years. Despite the progress, CTDE may suffer from the issue of Centralized-Decentralized Mismatch (CDM): the suboptimality of one agent's policy can exacerbate policy learning of other agents through the centralized joint critic. In contrast to centralized learning, the cooperative model that most closely resembles the way humans cooperate in nature is fully decentralized, i.e. Independent Learning (IL). However, there are still two issues that need to be addressed before agents coordinate through IL: (1) how agents are aware of the presence of other agents, and (2) how to coordinate with other agents to improve joint policy under IL. In this paper, we propose an inference-based coordinated MARL method: Deep Motor System (DMS). DMS first presents the idea of individual intention inference where agents are allowed to disentangle other agents from their environment. Secondly, causal inference was introduced to enhance coordination by reasoning each agent's effect on others' behavior. The proposed model was extensively experimented on a series of Multi-Agent MuJoCo and StarCraftII tasks. Results show that the proposed method outperforms independent learning algorithms and the coordination behavior among agents can be learned even without the CTDE paradigm compared to the state-of-the-art baselines including IPPO and HAPPO.

To meet the rising demand for software customization, highly configurable software systems play key roles in practice. Combinatorial interaction testing (CIT) is recognized as an effective approach for testing such systems. For CIT, the most important problem is constrained covering array generation (CCAG), which aims to construct a minimum-sized t-wise covering array (CA), where t denotes testing strength. Compared to pairwise testing (i.e., 2-wise CIT) that is a widely-used CIT technique, 3-wise CIT can discover more faults and bring more benefit in real-world applications. However, current state-of-the-art CCAG algorithms suffer from the severe scalability challenge for 3-wise CIT, which renders them ineffective in building 3-wise CAs for highly configurable systems. In this work, we perform an empirical study on various practical, highly configurable systems to present that it is promising to build 3-wise CA through extending 2-wise CA. Inspired by this, we propose ScalableCA, a novel and scalable algorithm that can effectively alleviate the scalability challenge for 3-wise CIT. Further, ScalableCA introduces three new and effective techniques, including fast invalidity detection, uncovering-guided sampling, and remainder-aware local search, to enhance its performance. Our experiments on extensive real-world, highly configurable systems show that, compared to current state-of-the-art algorithms, ScalableCA requires one to two orders of magnitude less running time to build 3-wise CA of 38.9% smaller size in average for large-scale instances. Our results indicate that ScalableCA greatly advances the state of the art in 3-wise CIT.

The increasing demand for software customization has led to the development of highly configurable systems. Combinatorial interaction testing (CIT) is an effective method for testing these types of systems. The ultimate goal of CIT is to generate a test suite of acceptable size, called a t-wise covering array (CA), where t is the testing strength. Pairwise testing (i.e., CIT with t=2) is recognized to be the most widely-used CIT technique and has strong fault detection capability. In pairwise testing, the most important problem is pairwise CA generation (PCAG), which is to generate a pairwise CA (PCA) of minimum size. However, existing state-of-the-art PCAG algorithms suffer from the severe scalability challenge; that is, they cannot tackle large-scale PCAG instances effectively, resulting in PCAs of large sizes. To alleviate this challenge, in this paper we propose CAmpactor, a novel and effective local search algorithm for compacting given PCAs into smaller sizes. Extensive experiments on a large number of real-world, public PCAG instances show that the sizes of CAmpactor's generated PCAs are around 45% smaller than the sizes of PCAs constructed by existing state-of-the-art PCAG algorithms, indicating its superiority. Also, our evaluation confirms the generality of CAmpactor, since CAmpactor can reduce the sizes of PCAs generated by a variety of PCAG algorithms.

Thanks to Network Function Virtualization (NFV), Internet Service Providers (ISPs) can improve network resource utilization with significantly reduced capital and operational expenditures. To dig deeper into the potential of NFV, an important challenge is the resource allocation problem in NFV (NFV-RA), which can be divided into three stages: VNFs chain composition, VNF forwarding graph embedding, and VNFs scheduling. The key to the NFV-RA problem is to design an effective and coordinated resource allocation algorithm for the three stages. Besides, the NFV-RA problem has been proved to be NP-Hard, and thus most existing approaches focus on heuristic and meta-heuristic algorithms. In this paper, we propose an NFV online coordinated resource allocation framework (OCRA) that completes the three stages simultaneously in a coordinated manner by combining parallel Multi-Agent Deep Reinforcement Learning with novel neural networks and RL training techniques. The extensive experimental results show that compared with the state-of-the-art solutions, OCRA is highly-efficient in terms of time, with up to 50% and 10.8% improvement on resource overhead and acceptance ratio, respectively.

Dynamic network representation learning has recently attracted increasing attention because real-world networks evolve over time, that is nodes and edges join or leave the networks over time. Different from static networks, the representation learning of dynamic networks should not only consider how to capture the structural information of network snapshots, but also consider how to capture the temporal dynamic information of network structure evolution from the network snapshot sequence. From the existing work on dynamic network representation, there are two main problems: (1) A significant number of methods target dynamic networks, which only allow nodes to increase over time, not decrease, which reduces the applicability of such methods to real-world networks. (2) At present, most network-embedding methods, especially dynamic network representation learning approaches, use Euclidean embedding space. However, the network itself is geometrically non-Euclidean, which leads to geometric inconsistencies between the embedded space and the underlying space of the network, which can affect the performance of the model. In order to solve the above two problems, we propose a geometry-based dynamic network learning framework, namely DyLFG. Our proposed framework targets dynamic networks, which allow nodes and edges to join or exit the network over time. In order to extract the structural information of network snapshots, we designed a new hyperbolic geometry processing layer, which is different from the previous literature. In order to deal with the temporal dynamics of the network snapshot sequence, we propose a gated recurrent unit (GRU) module based on Ricci curvature, that is the RGRU. In the proposed framework, we used a temporal attention layer and the RGRU to evolve the neural network weight matrix to capture temporal dynamics in the network snapshot sequence. The experimental results showed that our model outperformed the baseline approaches on the baseline datasets.

Graphs, as a non-linear data structure, are ubiquitous in practice, and efficient graph analysis can benefit important information retrieval applications in the era of big data. Currently, one of the fundamental graph mining problems is graph embedding, which aims to represent the graph as a low-dimensional feature vector with the content and structural information in the graph preserved. Although the graph embedding technique has evolved considerably, traditional methods mainly focus on node pairwise relationship in graphs, which makes the representational power of such schemes limited. Recently, a number of works have explored the simplicial complexes, which describe the higher-order interactions between nodes in the graphs, and further proposed several Graph Neural Network (GNN) algorithms based on simplicial complexes. However, these GNN approaches are highly inefficient in terms of running time and space, due to massive parameter learning. In this paper, we propose a simple and speedy graph embedding algorithm dubbed SCHash. Through adopting the Locality Sensitive Hashing (LSH) technique, SCHash captures the higher-order information derived from the simplicial complex in the GNN framework, and it can achieve a good balance between accuracy and efficiency. Our extensive experiments clearly show that, in terms of accuracy, the performance of our proposed SCHash algorithm is comparable to that of state-of-the-art GNN algorithms; also, SCHash achieves higher accuracy than the existing LSH algorithms. In terms of efficiency, SCHash runs faster than GNN algorithms by 2 ~ 4 orders of magnitude, and is more efficient than the existing LSH algorithms.