Query：
学者姓名：郭昆
Refining：
Year
Type
Indexed by
Source
Complex
Co-
Language
Clean All
Abstract ：
Link prediction in open knowledge graphs (OpenKGs) is crucial for applications like question answering and recommendation systems. Existing OpenKG models leverage the semantic information of noun phrases (NPs) to enhance the performance in the link prediction task. However, these models only extract superficial semantic information from NPs, ignoring the fact that an NP possesses diverse semantics. Furthermore, these models have not fully exploited the semantic information of the relation phrases (RPs). To address these issues, we propose a model for link prediction called Open Knowledge Graph Link Prediction with Semantic -Aware Embedding (SeAE). First, we develop an adaptive disentanglement embedding (ADE) mechanism to learn the intrinsically abundant semantics of NPs. The ADE mechanism can adaptively calculate the embedding segmentation number according to the dataset and has an ingenious method for updating embeddings. Second, we integrate the attention mechanism into the GRU encoder to obtain the distribution of importance inside RP, facilitating a more comprehensive capture of the RP's semantic information and enhancing the model's interpretability. Finally, we design a relation gate, which extracts the RP semantic features of tail NP from the shared edge. This gate realizes the relation constraints on entities while enhancing the interaction between entities and relations. Extensive experiments on four benchmarks demonstrate that SeAE outperforms the state-of-the-art models, resulting in improvements of approximately 5.4% and 7.4% in MRR on ReVerb45K and ReVerb45KF datasets respectively.
Keyword ：
Attention mechanism Attention mechanism Knowledge graph embedding Knowledge graph embedding Link prediction Link prediction Open knowledge graph Open knowledge graph Semantic-aware Semantic-aware
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Wang, Jingbin , Huang, Hao , Wu, Yuwei et al. Open Knowledge Graph Link Prediction with Semantic-Aware Embedding [J]. | EXPERT SYSTEMS WITH APPLICATIONS , 2024 , 249 . |
| MLA | Wang, Jingbin et al. "Open Knowledge Graph Link Prediction with Semantic-Aware Embedding" . | EXPERT SYSTEMS WITH APPLICATIONS 249 (2024) . |
| APA | Wang, Jingbin , Huang, Hao , Wu, Yuwei , Zhang, Fuyuan , Zhang, Sirui , Guo, Kun . Open Knowledge Graph Link Prediction with Semantic-Aware Embedding . | EXPERT SYSTEMS WITH APPLICATIONS , 2024 , 249 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Multi-label propagation algorithms (MLPAs) aim to find vertex communities in a complex network or a cloud system by propagating and updating vertex labels, which have been widely applied in customer recommendation, protein molecule discovery, and criminal tracking. As more and more people are concerned about the leakage of their sensitive information, detecting communities without disclosing personal privacy has become a hot topic in complex network analysis. The existing anonymization-based community detection methods have to modify the network structure to protect the sensitive vertices or links, which complicates the recognition of true communities and incurs substantial accuracy loss. In this article, we first propose a federated graph learning model (FGLM) for distributed privacy-preserving network data mining. Second, a federated MLPA for distributed and attributed networks is implemented by adapting a standalone MLPA to FGLM to verify the model's effectiveness. We develop a label perturbation strategy to conceal vertex degrees in distributed label updating and employ a homomorphic encryption system to protect label weights exchanged between the participants. The experiments on real-world and synthetic datasets demonstrate that the new algorithm achieves zero accuracy loss and more than 200% higher accuracy than the simple distributed MLPA without federated learning.
Keyword ：
Community detection Community detection Data privacy Data privacy Federated learning Federated learning Graph learning Graph learning Homomorphic encryption Homomorphic encryption Information integrity Information integrity Multi-label propagation Multi-label propagation Perturbation methods Perturbation methods Privacy Privacy Privacy-preserving data mining Privacy-preserving data mining Social networking (online) Social networking (online)
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Guo, Kun , Chen, Dangrun , Huang, Qingqing et al. Privacy-Preserving Multi-Label Propagation Based on Federated Learning [J]. | IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING , 2024 , 11 (1) : 886-899 . |
| MLA | Guo, Kun et al. "Privacy-Preserving Multi-Label Propagation Based on Federated Learning" . | IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING 11 . 1 (2024) : 886-899 . |
| APA | Guo, Kun , Chen, Dangrun , Huang, Qingqing , Li, Fuan , Guo, Chen , Wu, Duanji et al. Privacy-Preserving Multi-Label Propagation Based on Federated Learning . | IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING , 2024 , 11 (1) , 886-899 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Various datacenter network (DCN) load balancing schemes have been proposed in the past decade. Unfortunately, most of these solutions designed for lossy DCNs do not work well for Priority Flow Control (PFC) enabled lossless DCNs, primarily due to the reason that the individual congestion signals used in these solutions, e.g., link load, queue length, Round Trip Time (RTT) and Explicit Congestion Notification (ECN), may not be able to correctly or timely reflect the hop-by-hop PFC pausing. This paper first reveals the above problems via extensive experiments, and then based on the insights learned, we present Proteus, a PFC-aware load balancing scheme that is resilient to PFC pausing by exploring a combination of multi-level congestion signals. At its heart, Proteus leverages RTT-level signals (i.e., RTT and link utilization) to detect path status for initial routing decision, and exploits sub-RTT level signal (i.e., cumulative sojourn time) to reflect instantaneous PFC pausing and make timely rerouting choices based on the idea of better-late-than-never. We have implemented Proteus in the hardware programmable switch. Our testbed experiments as well as large-scale simulations show that Proteus can effectively handle PFC pausing under realistic workloads and achieve up to 35%, 31%, 28%, 22% and 46%, 42%, 34%, 29% better average FCT and 99(th) percentile FCT than CONGA, DRILL, Hermes and MP-RDMA, respectively.
Keyword ：
Computer science Computer science Datacenter Datacenter Delays Delays load balancing load balancing Load management Load management Load modeling Load modeling lossless networks lossless networks Receivers Receivers Switches Switches Transport protocols Transport protocols
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Hu, Jinbin , Zeng, Chaoliang , Wang, Zilong et al. Load Balancing With Multi-Level Signals for Lossless Datacenter Networks [J]. | IEEE-ACM TRANSACTIONS ON NETWORKING , 2024 , 32 (3) : 2736-2748 . |
| MLA | Hu, Jinbin et al. "Load Balancing With Multi-Level Signals for Lossless Datacenter Networks" . | IEEE-ACM TRANSACTIONS ON NETWORKING 32 . 3 (2024) : 2736-2748 . |
| APA | Hu, Jinbin , Zeng, Chaoliang , Wang, Zilong , Zhang, Junxue , Guo, Kun , Xu, Hong et al. Load Balancing With Multi-Level Signals for Lossless Datacenter Networks . | IEEE-ACM TRANSACTIONS ON NETWORKING , 2024 , 32 (3) , 2736-2748 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Graph neural networks have shown excellent performance in many fields owing to their powerful processing ability of graph data. In recent years, federated graph neural network has become a reasonable solution due to the enactment of privacy-related regulations. However, frequent communication between the coordinator and participants in federated graph neural network results in longer model training time and consumes many communication resources. To address this challenge, in this paper, we propose a novel semi-asynchronous federated graph learning communication protocol that simultaneously alleviates the negative impact of stragglers(slow participants) and accelerate the training process in the unsupervised federated graph neural network scenario. First, the weighted enforced synchronization strategy is intended to preserve the information carried by stragglers while preventing their stale models from harming the global model update. Second, the adaptive local update strategy is developed to make the local model of the participant with poor computing performance as close as possible to the global model. Experiments combine federated learning with graph contrastive learning. The results demonstrate that our proposed protocol outperforms the existing protocols in real-world networks. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Keyword ：
Graph neural networks Graph neural networks Internet protocols Internet protocols Learning systems Learning systems
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Liao, Yuanming , Wu, Duanji , Lin, Pengyu et al. Accelerating Unsupervised Federated Graph Neural Networks via Semi-asynchronous Communication [C] . 2024 : 378-392 . |
| MLA | Liao, Yuanming et al. "Accelerating Unsupervised Federated Graph Neural Networks via Semi-asynchronous Communication" . (2024) : 378-392 . |
| APA | Liao, Yuanming , Wu, Duanji , Lin, Pengyu , Guo, Kun . Accelerating Unsupervised Federated Graph Neural Networks via Semi-asynchronous Communication . (2024) : 378-392 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Community detection is essential for identifying cohesive groups in complex networks. Artificial benchmarks are critical for evaluating community detection algorithms, offering controlled environments with known community structures. However, existing benchmarks mainly focus on homogeneous networks and overlook the unique characteristics of heterogeneous networks. This paper proposes a novel artificial benchmark, called ABCD-HN (Artificial Network Benchmark for Community Detection on Heterogeneous Networks), for community detection in heterogeneous networks. This benchmark enables the generation of artificial heterogeneous networks with controllable community quantity, node quantity, and community complexity. Additionally, an evaluation framework for artificial heterogeneous networks is proposed to assess their effectiveness. Experimental results demonstrate the effectiveness and usability of ABCD-HN as a benchmark for artificial heterogeneous networks. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Keyword ：
Complex networks Complex networks Heterogeneous networks Heterogeneous networks Population dynamics Population dynamics
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Liu, Junjie , Guo, Kun , Wu, Ling . ABCD-HN: An Artificial Network Benchmark for Community Detection on Heterogeneous Networks [C] . 2024 : 182-194 . |
| MLA | Liu, Junjie et al. "ABCD-HN: An Artificial Network Benchmark for Community Detection on Heterogeneous Networks" . (2024) : 182-194 . |
| APA | Liu, Junjie , Guo, Kun , Wu, Ling . ABCD-HN: An Artificial Network Benchmark for Community Detection on Heterogeneous Networks . (2024) : 182-194 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
In order to address the problem of reconstruction and retraining time overhead in representation learning processing dynamic networks, this paper proposes an incremental inductive dynamic network community detection algorithm (IINDCD). First, the algorithm uses an attention mechanism to capture node neighborhood information and learn node representations by neighborhood aggregation induction while enhancing low-order structural representations. Second, the design uses random walking to capture high-order information and use it to construct node initial features for input into the attentional autoencoder, which effectively fuses high- and low-order structural features. Finally, the algorithm introduces the ideas of incremental update and model reuse for dynamic representation learning, constructs incremental node sets for updating the model, reduces training overhead, and quickly obtains node representation vectors for new moments of the network, then completing dynamic network community detection. IINDCD runs without reconstruction and with low retraining overhead. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Keyword ：
Learning algorithms Learning algorithms Learning systems Learning systems Population dynamics Population dynamics
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Wu, Ling , Zhuang, Jiangming , Guo, Kun . Incremental Inductive Dynamic Network Community Detection [C] . 2024 : 93-107 . |
| MLA | Wu, Ling et al. "Incremental Inductive Dynamic Network Community Detection" . (2024) : 93-107 . |
| APA | Wu, Ling , Zhuang, Jiangming , Guo, Kun . Incremental Inductive Dynamic Network Community Detection . (2024) : 93-107 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Community detection on attributed networks is a method to discover community structures within attributed networks. By applying community detection on attribute networks, we can better understand the relationships between nodes in real-world networks. However, current algorithms for community detection on attribute networks rely on hyper-parameters, and it is difficult to obtain an ideal result when facing networks with inconsistent attributes and topology. Consequently, we propose an Unsupervised Multi-population Evolutionary Algorithm (UMEA) for community detection in attributed networks. This algorithm adds edges between nodes based on attribute similarity, allowing it to combine attribute information during the process of community detection. In addition, this algorithm determines the optimal number of added edges autonomously through communication and learning between multiple populations. Furthermore, we propose a series of strategies to accelerate population convergence for the locus-based encoding. Experiments have demonstrated that our algorithm outperforms the benchmark algorithms in both real and artificial networks. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Keyword ：
Data mining Data mining Evolutionary algorithms Evolutionary algorithms Population dynamics Population dynamics
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Wu, Junjie , Wu, Lin , Guo, Kun . Unsupervised Multi-population Evolutionary Algorithm for Community Detection in Attributed Networks [C] . 2024 : 152-166 . |
| MLA | Wu, Junjie et al. "Unsupervised Multi-population Evolutionary Algorithm for Community Detection in Attributed Networks" . (2024) : 152-166 . |
| APA | Wu, Junjie , Wu, Lin , Guo, Kun . Unsupervised Multi-population Evolutionary Algorithm for Community Detection in Attributed Networks . (2024) : 152-166 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Discovering communities in attributed networks is an important research topic in complex network analysis. Community detection based on multi-objective evolutionary computing (MOEA) models community detection as a multi-objective optimization problem and searches the optimal solutions by simulating the evolution of a biological population. However, the existing multi-objective evolutionary algorithms for community detection faces two challenges: their encoding schemes are designed based on network topology and neglects the information in node attributes; and they are easy to fall into local optimum. In this article, we propose a community detection algorithm empowered by multi-objective evolutionary computing, named ECEVO-MOEA, which conducts edge closeness encoding and embedding vector optimization alternately. On the one hand, the evolution of a biological population is completed by employing a new edge closeness encoding scheme and multiple attribute-aware objective functions. On the other hand, the update of embedding vectors is used to calculate similarity matrix and communities to improve solution quality, avoiding it from early convergence. Experiments on real networks demonstrate that ECEVO-MOEA achieves higher accuracy than the baseline algorithms.
Keyword ：
Community networks Community networks Complex networks Complex networks Detection algorithms Detection algorithms Encoding Encoding Evolutionary computation Evolutionary computation Image edge detection Image edge detection Pareto optimization Pareto optimization Search problems Search problems Social factors Social factors Statistics Statistics
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Guo, Kun , Chen, Zhanhong , Yu, Zhiyong et al. Evolutionary Computing Empowered Community Detection in Attributed Networks [J]. | IEEE COMMUNICATIONS MAGAZINE , 2024 , 62 (5) : 22-26 . |
| MLA | Guo, Kun et al. "Evolutionary Computing Empowered Community Detection in Attributed Networks" . | IEEE COMMUNICATIONS MAGAZINE 62 . 5 (2024) : 22-26 . |
| APA | Guo, Kun , Chen, Zhanhong , Yu, Zhiyong , Chen, Kai , Guo, Wenzhong . Evolutionary Computing Empowered Community Detection in Attributed Networks . | IEEE COMMUNICATIONS MAGAZINE , 2024 , 62 (5) , 22-26 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
A clear game process helps to track community generation and evolution, so non-cooperative and cooperative games are applied to community detection. However, non-cooperative games focus on the competition between nodes, disregarding their cooperation. Solely considering individual perspectives often results in insufficient precision. Cooperative games consider the interests of both coalitions and individual. Nevertheless, involving a large number of participants in cooperative games can lead to high computational complexity and slow convergence. In this study, a fast community detection model called FCDG is proposed. It combines non-cooperative and cooperative games by exploring candidate communities and optimizing community merging. Firstly, a intimate core group identification strategy based on node mutual intimacy is designed to accelerate the convergence of candidate community detection using non-cooperative games and maximize individual benefits. Secondly, building upon of candidate community, a candidate community merging approach based on cooperative games is devised to achieve community optimal solution. The performance of FCDG is evaluated on both real-world and synthetic datasets. Experimental results demonstrate that FCDG effectively discovers community structure with higher accuracy and robustness compared to other baseline algorithms. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Keyword ：
Complex networks Complex networks Game theory Game theory Lead compounds Lead compounds Merging Merging Population dynamics Population dynamics
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Wu, Ling , Yuan, Mao , Guo, Kun . Fast Community Detection Based on Integration of Non-cooperative and Cooperative Game [C] . 2024 : 276-286 . |
| MLA | Wu, Ling et al. "Fast Community Detection Based on Integration of Non-cooperative and Cooperative Game" . (2024) : 276-286 . |
| APA | Wu, Ling , Yuan, Mao , Guo, Kun . Fast Community Detection Based on Integration of Non-cooperative and Cooperative Game . (2024) : 276-286 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Community detection is widely used in network analysis, which seeks to divide network nodes into distinct communities based on the topology structure and attribute information of the network. Due to its interpretability, nonnegative matrix factorization becomes an essential method for community detection. However, it decomposes the adjacency matrix and attribute matrix separately, which do not tightly incorporate topology and attributes. And in the problem of division inconsistency based on topology and attributes caused by the mismatch between the topology similarity and attribute similarity of paired nodes, it ignores the difference in the matching degree of each attribute and each node. In this paper, we propose a nonnegative matrix factorization algorithm for community detection (MTACD) based on the matching degree between topology and attribute. First, we employ an attribute embedding mechanism to enhance the node-attribute relationship. Second, we design an attribute matching degree and a node topology-and-attribute matching degree in order to resolve the mismatch between topology and attribute similarity. Experiments on both real-world and synthetic networks demonstrate the effectiveness of our algorithm. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
Keyword ：
Matrix algebra Matrix algebra Matrix factorization Matrix factorization Population dynamics Population dynamics Topology Topology
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Zeng, Ruolan , Liu, Zhanghui , Guo, Kun . Nonnegative Matrix Factorization Based on Topology-and-Attribute-Matching Degree for Community Detection [C] . 2024 : 137-151 . |
| MLA | Zeng, Ruolan et al. "Nonnegative Matrix Factorization Based on Topology-and-Attribute-Matching Degree for Community Detection" . (2024) : 137-151 . |
| APA | Zeng, Ruolan , Liu, Zhanghui , Guo, Kun . Nonnegative Matrix Factorization Based on Topology-and-Attribute-Matching Degree for Community Detection . (2024) : 137-151 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Export
| Results： |
Selected to |
| Format： |
