共查询到20条相似文献，搜索用时 15 毫秒
1.
《Information processing & management》2022,59(3):102916
Graph convolutional network (GCN) is a powerful tool to process the graph data and has achieved satisfactory performance in the task of node classification. In general, GCN uses a fixed graph to guide the graph convolutional operation. However, the fixed graph from the original feature space may contain noises or outliers, which may degrade the effectiveness of GCN. To address this issue, in this paper, we propose a robust graph learning convolutional network (RGLCN). Specifically, we design a robust graph learning model based on the sparse constraint and strong connectivity constraint to achieve the smoothness of the graph learning. In addition, we introduce graph learning model into GCN to explore the representative information, aiming to learning a high-quality graph for the downstream task. Experiments on citation network datasets show that the proposed RGLCN outperforms the existing comparison methods with respect to the task of node classification. 相似文献
2.
Roland R. Mielke 《Journal of The Franklin Institute》1978,305(3):143-159
Topological analysis of the signal flow graph associated with the hybrid system of equations for a linear active or passive electrical network for which the element admittance matrix exists and is diagonal is considered. First, the term cancellation which occurs in Mason's topological formulas is investigated. Necessary and sufficient conditions on the signal flow graph topology such that a term in the expansion of the graph determinant and cofactors either cancels out with another term in the expansion or does not cancel are established. Properties of the associated network which result in non-cancelling terms are given and the number of non-cancelling terms is determined. Second, new signal flow graph topological formulas for the graph determinant and cofactors are proven. These formulas are such that no term cancellation occurs and are readily adaptable to computer implementation. In addition, the number of terms in these formulas is independent of the network tree used to formulate the signal flow graph. Examples are given to illustrate the new formulas. 相似文献
3.
网络结构与创新扩散研究 总被引：1，自引：0，他引：1
根据WS小世界模型的思想，构建了从规则网络到随机网络的一系列扩散网络图，通过虚拟采纳个体的决策过程，研究了网络结构与性质对创新微观采纳和宏观扩散的影响。不同于已往的研究，本文将创新采纳与创新扩散统一起来，运用复杂网络的方法，研究了网络结构与创新扩散之间的动态相互关系。数值模拟结果表明：存在介于规则网络和随机网络之间的小世界扩散网络；外部因素和内部因素共同决定一个成功的扩散过程；网络的簇系数决定扩散的最终水平，而网络平均距离决定扩散速度；网络个体间的异质性程度越大，越不利于创新扩散。 相似文献
4.
【目的/意义】随着网络和多媒体技术的迅速发展,知识虚拟社区作为一种以实现教育和学习为目的的网络
学习共同体,亦成为学习者进行知识共享和经验交流的重要平台。对教育虚拟社区中知识传播的网络结构研究能
够帮助认识知识扩散模式,有利于虚拟社区中的知识组织和知识管理。【方法/过程】本文基于知识图谱理论以
MOOC（慕课）为例构建知识传播社区图谱并分析其网络结构,使用社会网络分析方法分别从宏观和微观剖析知识
传播社区中以学生学习课程构成的网络的互动结构特征和互动关系特征,挖掘知识传播模式。【结果/结论】研究结
果揭示虚拟教育学习社区各节点的关联关系以及路径距离。学生学习课程网络中的节点连接紧密,形成的社区较
明显;具有较高中心度和PageRank值的课程更受学生欢迎;具有较高中心度和PageRank值的学生更积极参与课程
互动。研究结果为在线教育知识传播社区用户管理和知识管理提供依据和方法支撑。【创新/局限】论文从理论层
面论述了知识图谱理论构建教育虚拟社区知识转移及知识传播网络,通过可视化图谱呈现知识社区网络结构;
MOOC作为数据源分析教育虚拟社区网络结构,且采集的样本数据量还有待增加,具有一定局限性。 相似文献
5.
In this paper, an interventional bipartite consensus problem is considered for a high-order multi-agent system with unknown disturbance dynamics. The interactions among the agents are cooperative and competitive simultaneously and thus the interaction network (just called coopetition network in sequel for simplicity) is conveniently modeled by a signed graph. When the coopetition network is structurally balanced, all the agents are split into two competitive subgroups. An exogenous system (called leader for simplicity) is introduced to intervene the two competitive subgroups such that they can reach a bipartite consensus. The unknown disturbance dynamics are assumed to have linear parametric models. With the help of the notation of a disagreement state variable, decentralized adaptive laws are proposed to estimate the unknown disturbances and a dynamic output-feedback consensus control is designed for each agent in a fully distributed fashion, respectively. The controller design guarantees that the state matrix of the closed-loop system can be an arbitrary predefined Hurwitz matrix. Under the assumption that the coopetition network is structurally balanced and the leader is a root of the spanning tree in an augmented graph, the bipartite consensus and the parameter estimation are analyzed by invoking a common Lyapunov function method when the coopetition network is time-varying according to a piecewise constant switching signal. Finally, simulation results are given to demonstrate the effectiveness of the proposed control strategy. 相似文献
6.
《Information processing & management》2022,59(4):102953
Aspect-based sentiment analysis technologies may be a very practical methodology for securities trading, commodity sales, movie rating websites, etc. Most recent studies adopt the recurrent neural network or attention-based neural network methods to infer aspect sentiment using opinion context terms and sentence dependency trees. However, due to a sentence often having multiple aspects sentiment representation, these models are hard to achieve satisfactory classification results. In this paper, we discuss these problems by encoding sentence syntax tree, words relations and opinion dictionary information in a unified framework. We called this method heterogeneous graph neural networks (Hete_GNNs). Firstly, we adopt the interactive aspect words and contexts to encode the sentence sequence information for parameter sharing. Then, we utilized a novel heterogeneous graph neural network for encoding these sentences’ syntax dependency tree, prior sentiment dictionary, and some part-of-speech tagging information for sentiment prediction. We perform the Hete_GNNs sentiment judgment and report the experiments on five domain datasets, and the results confirm that the heterogeneous context information can be better captured with heterogeneous graph neural networks. The improvement of the proposed method is demonstrated by aspect sentiment classification task comparison. 相似文献
7.
《Information processing & management》2023,60(2):103207
Visual Question Answering (VQA) requires reasoning about the visually-grounded relations in the image and question context. A crucial aspect of solving complex questions is reliable multi-hop reasoning, i.e., dynamically learning the interplay between visual entities in each step. In this paper, we investigate the potential of the reasoning graph network on multi-hop reasoning questions, especially over 3 “hops.” We call this model QMRGT: A Question-Guided Multi-hop Reasoning Graph Network. It constructs a cross-modal interaction module (CIM) and a multi-hop reasoning graph network (MRGT) and infers an answer by dynamically updating the inter-associated instruction between two modalities. Our graph reasoning module can apply to any multi-modal model. The experiments on VQA 2.0 and GQA (in fully supervised and O.O.D settings) datasets show that both QMRGT and pre-training V&L models+MRGT lead to improvement on visual question answering tasks. Graph-based multi-hop reasoning provides an effective signal for the visual question answering challenge, both for the O.O.D and high-level reasoning questions. 相似文献
8.
《Information processing & management》2023,60(4):103377
The recent boom in online courses has necessitated personalized online course recommendation. Modelling the learning sequences of users is key for course recommendation because the sequences contain the dynamic learning interests of the users. However, current course recommendation methods ignore heterogeneous course information and collective sequential dependency between courses when modelling the learning sequences. We thus propose a novel online course recommendation method based on knowledge graph and deep learning which models course information via a course knowledge graph and represents courses using TransD. It then develops a bidirectional long short-term memory network, convolutional neural network, and multi-layer perceptron for learning sequence modelling and course recommendation. A public dataset called MOOCCube was used to evaluate the proposed method. Experimental results show that: (1) employing the course knowledge graph in learning sequence modelling improves averagely the performance of our method by 13.658%, 16.42%, and 15.39% in terms of HR@K, MRR@K, and NDCG@K; (2) modelling the collective sequential dependency improves averagely the performance by 4.11%, 6.37%, and 5.47% in terms of the above metrics; and (3) our method outperforms popular methods with the course knowledge graph in most cases. 相似文献
9.
《Information processing & management》2023,60(4):103387
Precise prediction of Multivariate Time Series (MTS) has been playing a pivotal role in numerous kinds of applications. Existing works have made significant efforts to capture temporal tendency and periodical patterns, but they always ignore abrupt variations and heterogeneous/spatial associations of sensory data. In this paper, we develop a dual normalization (dual-norm) based dynamic graph diffusion network (DNGDN) to capture hidden intricate correlations of MTS data for temporal prediction. Specifically, we design time series decomposition and dual-norm mechanism to learn the latent dependencies and alleviate the adverse effect of abnormal MTS data. Furthermore, a dynamic graph diffusion network is adopted for adaptively exploring the spatial correlations among variables. Extensive experiments are performed on 3 real world experimental datasets with 8 representative baselines for temporal prediction. The performances of DNGDN outperforms all baselines with at least 4% lower MAPE over all datasets. 相似文献
10.
Many science and engineering problems can be represented by a network, a generalization of which is a graph. Examples of the problems that can be represented by a graph include: cyclic sequential circuit, organic molecule structures, mechanical structures, etc. The most fundamental issue with these problems (e.g., designing a molecule structure) is the identification of structure, which further reduces to be the identification of graph. The problem of the identification of graph is called graph isomorphism. The graph isomorphism problem is an NP problem according to the computational complexity theory. Numerous methods and algorithms have been proposed to solve this problem. Elsewhere we presented an approach called the eigensystem approach. This approach is based on a combination of eigenvalue and eigenvector which are further associated with the adjacency matrix. The eigensystem approach has been shown to be very effective but requires that a graph must contain at least one distinct eigenvalue. The adjacency matrix is not shown sufficiently to meet this requirement. In this paper, we propose a new matrix called adjusted adjacency matrix that meets this requirement. We show that the eigensystem approach based on the adjusted adjacency matrix is not only effective but also more efficient than that based on the adjacency matrix. 相似文献
11.
《Information processing & management》2023,60(5):103439
Graph neural networks have been frequently applied in recommender systems due to their powerful representation abilities for irregular data. However, these methods still suffer from the difficulties such as the inflexible graph structure, sparse and highly imbalanced data, and relatively shallow networks, limiting rate prediction ability for recommendations. This paper presents a novel deep dynamic graph attention framework based on influence and preference relationship reconstruction (DGA-IPR) for recommender systems to learn optimal latent representations of users and items. The entire framework involves a user branch and an item branch. An influence-based dynamic graph attention (IDGA) module, a preference-based dynamic graph attention (PDGA) module, and an adaptive fine feature extraction (AFFE) module are respectively constructed for each branch. Concretely, the first two attention modules concentrate on reconstructing influence and preference relationship graphs, breaking imbalanced and fixed constraints of graph structures. Then a deep feature aggregation block and an adaptive feature fusion operation are built, improving the network depth and capturing potential high-order information expressions. Besides, AFFE is designed to acquire finer latent features for users and items. The DGA-IPR architecture is formed by integrating IDGA, PDGA, and AFFE for users and items, respectively. Experiments reveal the superiority of DGA-IPR over existing recommendation models. 相似文献
12.
《Information processing & management》2022,59(3):102938
Previous studies on Course Recommendation (CR) mainly focus on investigating the sequential relationships among courses (RNN is applied) and fail to learn the similarity relationships among learners. Moreover, existing RNN-based methods can only model courses’ short-term sequential patterns due to the inherent shortcomings of RNNs. In light of the above issues, we develop a hyperedge-based graph neural network, namely HGNN, for CR. Specifically, (1) to model the relationships among learners, we treat learners (i.e., hyperedges) as the sets of courses in a hypergraph, and convert the task of learning learners’ representations to induce the embeddings for hyperedges, where a hyperedge-based graph attention network is further proposed. (2) To simultaneously consider courses’ long-term and short-term sequential relationships, we first construct a course sequential graph across learners, and learn courses’ representations via a modified graph attention network. Then, we feed the learned representations into a GRU-based sequence encoder to infer their short-term patterns, and deem the last hidden state as the learned sequence-level learner embedding. After that, we obtain the learners’ final representations by a product pooling operation to retain features from different latent spaces, and optimize a cross-entropy loss to make recommendations. To evaluate our proposed solution HGNN, we conduct extensive experiments on two real-world datasets, XuetangX and MovieLens. We conduct experiments on MovieLens to prove the extensibility of our solution on other collections. From the experimental results, we can find that HGNN evidently outperforms other recent CR methods on both datasets, achieving 11.96% on P@20, 16.01% on NDCG@20, and 27.62% on MRR@20 on XuetangX, demonstrating the effectiveness of studying CR in a hypergraph, and the importance of considering both long-term and short-term sequential patterns of courses. 相似文献
13.
In this paper, we investigate the output synchronization of networked SISO nonlinear systems that can be transformed into semi-strict feedback form. Due to parameter uncertainty, the agents have heterogeneous dynamics. Combined backstepping method together with graph theory, we construct an augmented Laplacian potential function for analysis and a distributed controller is designed recursively for each agent such that its output can be synchronized to its neighbors' outputs. The distributed controller of each agent has three parts: state feedback of itself, neighborhood information transmitted through the network and adaptive parameter updaters both for itself and its neighbors. Moreover, distributed tuning function is designed to minimize the order of the parameter updater. It is proved that when the undirected graph is connected, all agents’ outputs in the network can be synchronized, i.e., cooperative output synchronization of the network is realized. Simulation results are presented to verify the effectiveness of the proposed controllers. 相似文献
14.
排序是信息检索、数据挖掘以及社会网络分析的基础工作之一。 在线社交网络和社
会媒体的快速发展积累了大量的图数据——由表示实体的节点和表示实体间关系的连边构
成。 图数据中节点之间连接关系复杂, 通常缺少显式的全序结构, 使得图排序在图数据分析
中显得尤为重要。 图排序算法主要包括 2 大类, 面向节点中心度的图排序算法和面向节点集
合多样性的图排序算法。 与传统的图排序不同 , 多样性图排序考虑排序和聚类的融合, 体现
为节点集合对网络整体的覆盖程度。 近年来, 多样性图排序得到了广泛的关注, 取得了一系
列研究进展,研究成果成功应用到了搜索结果排序、文档自动摘要、信息推荐系统和影响最大
化等诸多场景中。 文章评述了多样性图排序的研究现状及主要进展, 将现有的多样性图排序
方法按照研究思路的不同分为边际效益最大化、竞争随机游走、聚类与排序互增强 3 类, 分别
评述了每类方法的优势和不足。 最后指出 , 设计有效的评价指标和标准测试集、克服多样性
图排序面临的精度和速度的矛盾等是多样性图排序未来的研究重点。 相似文献
15.
[目的/意义]探索基于知识图谱的网络社区学术资源深度聚合的理论和方法,为网络学术社区知识细粒度组织、知识服务实践提供思路引导和新视角。[方法/过程]首先梳理了知识图谱和学术资源聚合的研究进展,从价值需求主体的角度剖析网络社区学术资源聚合的应用价值;然后明确网络社区学术知识图谱的构建流程,构建出基于知识图谱的网络社区学术资源深度聚合框架,并介绍知识富关联关系提取方法;最后设计个体用户画像、语义智能检索、分面式导航三种应用模式。[结果/结论]知识图谱能够较好地应用于学术资源深度聚合,支持网络社区的高级知识服务应用,基于知识图谱的网络社区学术资源深度聚合框架对学术类网络社区平台优化资源配置、有效知识创新服务具有重要参考价值。 相似文献
16.
This paper discusses adaptive synchronization control for complex networks interacted in an undirected weighted graph, and aims to provide a novel and general approach for the design of distributed update laws for adaptively adjusting coupling weights. The proposed updating laws are very general in the sense that they encompass most weight update laws reported in the literature as special cases, and also provide new insights in the analysis of network system evolution and graph weight convergence. We show a rigorous proof for the synchronization stability of the overall complex network to a synchronized state, and demonstrate the convergence of adaptive weights for each edge to some bounded constants. A detailed comparison with available results is provided to elaborate the new features and advantages of the proposed adaptive strategies as compared with conventional adaptive laws. The effectiveness of the proposed approach is also validated by several typical simulations. 相似文献
17.
文章对中国知网近10年机构知识库相关文献进行定量分析,采用Ucinet、Pajek等可视化软件,结合社会网络分析等方法,绘制了作者、机构、期刊与关键词等网络知识图谱。基于知识图谱视角,更为直观地分析了国内机构知识库研究现状,较为准确地揭示了国内研究热点,对国内外机构知识库的研究现状和研究热点进行了对比和分析。 相似文献
18.
《Information processing & management》2023,60(3):103292
Dynamic link prediction is a critical task in network research that seeks to predict future network links based on the relative behavior of prior network changes. However, most existing methods overlook mutual interactions between neighbors and long-distance interactions and lack the interpretability of the model’s predictions. To tackle the above issues, in this paper, we propose a temporal group-aware graph diffusion network(TGGDN). First, we construct a group affinity matrix to describe mutual interactions between neighbors, i.e., group interactions. Then, we merge the group affinity matrix into the graph diffusion to form a group-aware graph diffusion, which simultaneously captures group interactions and long-distance interactions in dynamic networks. Additionally, we present a transformer block that models the temporal information of dynamic networks using self-attention, allowing the TGGDN to pay greater attention to task-related snapshots while also providing interpretability to better understand the network evolutionary patterns. We compare the proposed TGGDN with state-of-the-art methods on five different sizes of real-world datasets ranging from 1k to 20k nodes. Experimental results show that TGGDN achieves an average improvement of 8.3% and 3.8% in terms of ACC and AUC on all datasets, respectively, demonstrating the superiority of TGGDN in the dynamic link prediction task. 相似文献
19.
20.
用R语言分析关键词集共现网络研究 总被引：1，自引：0，他引：1
[目的/意义]提出关键词集的概念,探索R语言编程实现关键词集共现网络的创建和可视化,为进一步研究基于关键词集的数据挖掘和知识发现提供更为通用的途径和方法。[方法/过程]运用R语言编程技术及igraph等贡献包,自编了关键词集共现网络的创建和可视化函数,分析了图情学科领域的18种CSSCI源刊的载文数据。[结果/结论]计算了关键词集共现网络的中心性等特征参数,绘制了关键词集共现网络图。研究表明,关键词集共现网络揭示了关键词集的分布、聚类和关系特征,能更为直观的揭示分析对象的主题内容及其关联关系,其特征参数的构建及其表征等理论问题值得系统而深入的研究。 相似文献