基于Hypergraph的突发事件情景案例表示与检索方法研究

[研究目的]突发事件具有情景演化特征,突发事件案例不仅要对单个情景特征进行描述,也要描述不同情景之间的演化关系。[研究方法]在对突发事件情景演化机理和情景案例特征分析基础上,引入超图理论,将其应用于突发事件情景案例表示,建立了突发事件情景链超图模型,并基于该模型提出了包括情景链检索和情景特征检索的两阶段检索策略和相应的相似度计算方法。[研究结论]该模型能够有效描述突发事件情景案例的复杂演化特征,并能提高案例检索的质量和效果。

The Erdös-Faber-Lovász Conjecture for Gap-Restricted Hypergraphs

An edge coloring of hypergraph H is a function such that holds for any pair of intersecting edges . The minimum number of colors in edge colorings of H is called the chromatic index of H and is denoted by . Erdös, Faber and Lovász proposed a famous conjecture that holds for any loopless linear hypergraph H with n vertices. In this paper, we show that is true for gap-restricted hypergraphs. Our result extends a result of Alesandroni in 2021.

Some Ore-type Results for Matching and Perfect Matching in k-uniform Hypergraphs

Let SI and S2 be two （k- 1）-subsets in a k-uniform hypergraph H. We call S1 and S2 strongly or middle or weakly independent if H does not contain an edge e ∈ E（H） such that S1 ∩ e≠ 0 and S2 ∩ e ≠0 or e S1 ∪ S2 or e S1 ∪ S2, respectively. In this paper, we obtain the following results concerning these three independence. （1） For any n ≥ 2k2 - k and k ≥ 3, there exists an n-vertex k-uniform hypergraph, which has degree sum of any two strongly independent （k - 1）-sets equal to 2n - 4（k - 1）, contains no perfect matching; （2） Let d ≥ 1 be an integer and H be a k-uniform hypergraph of order n ≥ kd＋ （k- 2）k. If the degree sum of any two middle independent （k- 1）-subsets is larger than 2（d- 1）, then H contains a d-matching; （3） For all k ≥ 3 and sufficiently large n divisible by k, we completely determine the minimum degree sum of two weakly independent （k - 1）-subsets that ensures a perfect matching in a k-uniform hypergraph H of order n.

Diverse Deep Matrix Factorization With Hypergraph Regularization for Multi-View Data Representation

Deep matrix factorization(DMF)has been demonstrated to be a powerful tool to take in the complex hierarchical information of multi-view data(MDR).However,existing multiview DMF methods mainly explore the consistency of multi-view data,while neglecting the diversity among different views as well as the high-order relationships of data,resulting in the loss of valuable complementary information.In this paper,we design a hypergraph regularized diverse deep matrix factorization(HDDMF)model for multi-view data representation,to jointly utilize multi-view diversity and a high-order manifold in a multilayer factorization framework.A novel diversity enhancement term is designed to exploit the structural complementarity between different views of data.Hypergraph regularization is utilized to preserve the high-order geometry structure of data in each view.An efficient iterative optimization algorithm is developed to solve the proposed model with theoretical convergence analysis.Experimental results on five real-world data sets demonstrate that the proposed method significantly outperforms stateof-the-art multi-view learning approaches.

A Nonlinear Spatiotemporal Optimization Method of Hypergraph Convolution Networks for Traffic Prediction

Traffic prediction is a necessary function in intelligent transporta-tion systems to alleviate traffic congestion.Graph learning methods mainly focus on the spatiotemporal dimension,but ignore the nonlinear movement of traffic prediction and the high-order relationships among various kinds of road segments.There exist two issues:1)deep integration of the spatiotempo-ral information and 2)global spatial dependencies for structural properties.To address these issues,we propose a nonlinear spatiotemporal optimization method,which introduces hypergraph convolution networks(HGCN).The method utilizes the higher-order spatial features of the road network captured by HGCN,and dynamically integrates them with the historical data to weigh the influence of spatiotemporal dependencies.On this basis,an extended Kalman filter is used to improve the accuracy of traffic prediction.In this study,a set of experiments were conducted on the real-world dataset in Chengdu,China.The result showed that the proposed method is feasible and accurate by two different time steps.Especially at the 15-minute time step,compared with the second-best method,the proposed method achieved 3.0%,11.7%,and 9.0%improvements in RMSE,MAE,and MAPE,respectively.

Brain Functional Networks with Dynamic Hypergraph Manifold Regularization for Classification of End-Stage Renal Disease Associated with Mild Cognitive Impairment

The structure and function of brain networks have been altered in patients with end-stage renal disease(ESRD).Manifold regularization(MR)only considers the pairing relationship between two brain regions and cannot represent functional interactions or higher-order relationships between multiple brain regions.To solve this issue,we developed a method to construct a dynamic brain functional network(DBFN)based on dynamic hypergraph MR(DHMR)and applied it to the classification of ESRD associated with mild cognitive impairment(ESRDaMCI).The construction of DBFN with Pearson’s correlation(PC)was transformed into an optimization model.Node convolution and hyperedge convolution superposition were adopted to dynamically modify the hypergraph structure,and then got the dynamic hypergraph to form the manifold regular terms of the dynamic hypergraph.The DHMR and L_(1) norm regularization were introduced into the PC-based optimization model to obtain the final DHMR-based DBFN(DDBFN).Experiment results demonstrated the validity of the DDBFN method by comparing the classification results with several related brain functional network construction methods.Our work not only improves better classification performance but also reveals the discriminative regions of ESRDaMCI,providing a reference for clinical research and auxiliary diagnosis of concomitant cognitive impairments.

Hypergraph Regularized Deep Autoencoder for Unsupervised Unmixing Hyperspectral Images

Deep learning(DL)has shown its superior performance in dealing with various computer vision tasks in recent years.As a simple and effective DL model,autoencoder(AE)is popularly used to decompose hyperspectral images(HSIs)due to its powerful ability of feature extraction and data reconstruction.However,most existing AE-based unmixing algorithms usually ignore the spatial information of HSIs.To solve this problem,a hypergraph regularized deep autoencoder(HGAE)is proposed for unmixing.Firstly,the traditional AE architecture is specifically improved as an unsupervised unmixing framework.Secondly,hypergraph learning is employed to reformulate the loss function,which facilitates the expression of high-order similarity among locally neighboring pixels and promotes the consistency of their abundances.Moreover,L_(1/2)norm is further used to enhance abundances sparsity.Finally,the experiments on simulated data,real hyperspectral remote sensing images,and textile cloth images are used to verify that the proposed method can perform better than several state-of-the-art unmixing algorithms.

Load-Aware VM Migration Using Hypergraph Based CDB-LSTM

Live Virtual Machine(VM)migration is one of the foremost techniques for progressing Cloud Data Centers’(CDC)proficiency as it leads to better resource usage.The workload of CDC is often dynamic in nature,it is better to envisage the upcoming workload for early detection of overload status,underload status and to trigger the migration at an appropriate point wherein enough number of resources are available.Though various statistical and machine learning approaches are widely applied for resource usage prediction,they often failed to handle the increase of non-linear CDC data.To overcome this issue,a novel Hypergrah based Convolutional Deep Bi-Directional-Long Short Term Memory(CDB-LSTM)model is proposed.The CDB-LSTM adopts Helly property of Hypergraph and Savitzky–Golay(SG)filter to select informative samples and exclude noisy inference&outliers.The proposed approach optimizes resource usage prediction and reduces the number of migrations with minimal computa-tional complexity during live VM migration.Further,the proposed prediction approach implements the correlation co-efficient measure to select the appropriate destination server for VM migration.A Hypergraph based CDB-LSTM was vali-dated using Google cluster dataset and compared with state-of-the-art approaches in terms of various evaluation metrics.

k一致超图的α谱极值结果

设F是一个简单图,如果Berge F的每条边均由超边替换F中的边而得到,则Berge F为超图.超图G如果不包含子超图Berge F,则G是Berge F-free的.为此,基于Aα张量研究具有特殊结构的线性一致超图的谱-Turán-问题,分别证明了Berge C4-free和围长至少为5的线性一致超图的α谱极值.

基于非回溯矩阵中心性的超图可靠性研究

近年来,超图作为网络科学的一个研究热点,引起了广泛的关注。超图区别于传统图的结构特点在于它的超边可以同时连接多个节点,从而形成更为复杂和高阶的关系。在这样的网络结构中,有效地识别重要的节点和超边成为一个关键的挑战。特征向量中心性是一个常见的度量标准,但当网络中存在着极大度值的枢纽节点时,使用特征向量中心性度量方法会使结果表现出局域性,限制了该方法的应用场景。因此,将超图转化成对应的线图,在此基础上使用非回溯矩阵中心性这一方法,该方法在评估超边重要性时表现出更好的均匀性和区分度。此外,还探讨了特征向量中心性和非回溯矩阵中心性在超图中节点重要性评估上的应用。通过比较这两种方法,研究发现非回溯矩阵中心性在区分节点重要程度方面具有更明显的优势。研究不仅包括理论分析和模型构建,还包括对真实世界数据的实证。为了验证所提方法和结论,选取了6个真实世界超图作为实验对象。通过在这些超图上的应用,证明了非回溯矩阵中心性在识别重要节点和超边方面的有效性。研究为超图中关键元素的识别提供了一种新的视角和方法,对于理解和分析实际复杂网络系统,具有重要的理论和实践意义。

注意力感知的边−节点交换图神经网络模型

提出了一种注意力感知的边-节点交换图神经网络(attention aware edge-node exchange graph neural network,AENN)模型,在图结构化数据表示框架下,使用边-节点切换卷积的图神经网络算法进行图编码,用于半监督分类和回归分析。AENN是一种通用的图编码框架,用于将图节点和边嵌入一个统一的潜在特征空间。具体地,基于原始无向图,不断切换边与节点的卷积,并在卷积过程中通过注意力机制分配不同邻居的权重,从而实现特征传播。在3个数据集上的实验研究表明,所提方法较已有方法在半监督分类和回归分析中具有明显的性能提升。

基于超图嵌入和有限注意力的社会化推荐

近年来,社会化推荐作为推荐算法之一被广泛应用于各大平台.由于引入了用户的社交信息,社会化推荐可以较好地缓解数据稀疏问题.然而,大部分社会化推荐难以高效地从原始信息中提取用户的有效信息,导致引入社会信息的同时也会引入大量噪声.为了解决上述问题,本文提出了SRBHL(Social Recommendation Based on Hypergraph embedding and Limited attention)模型,通过超图嵌入模块提取用户的历史行为信息和社交信息,以缓解原始目标用户数据稀疏问题,并结合有限注意力模块来过滤原始信息的噪声,最后将得到的有效好友信息用于推荐.在Yelp-Urbana、Yelp-Phoenix和Epinions3个真实数据集上的实验结果表明SRBHL模型相比其他的推荐算法表现更出色.此外,本文还对SRBHL模型进行了鲁棒性分析,并给出了模型最优参数的取值范围.

基于超图Transformer的APT攻击威胁狩猎网络模型

针对物联网环境中高级持续性威胁(APT)具有隐蔽性强、持续时间长、更新迭代快等特点,传统被动检测模型难以对其进行有效搜寻的问题,提出了一种基于超图Transformer的APT攻击威胁狩猎(HTTN)模型,能够在时间跨度长、信息隐蔽复杂的物联网系统中快速定位和发现APT攻击痕迹。该模型首先将输入的网络威胁情报(CTI)日志图和物联网系统内核审计日志图编码为超图,经超图神经网络(HGNN)层计算日志图的全局信息和节点特征;然后由Transformer编码器提取超边位置特征;最后对超边进行匹配计算相似度分数,从而实现物联网系统网络环境下APT攻击的威胁狩猎。在物联网仿真环境下的实验结果表明,提出的HTTN模型与目前主流的图匹配神经网络相比均方误差降低约20%,Spearman等级相关系数提升约0.8%,匹配精度提升约1.2%。

基于BERT和超图对偶注意力网络的文本情感分析

针对网络短文本存在大量的噪声和缺乏上下文信息的问题,提出一种基于BERT和超图对偶注意力机制的文本情感分析模型。首先利用BERT预训练模型强大的表征学习能力,对情感文本进行动态特征提取;同时挖掘文本的上下文顺序信息、主题信息和语义依存信息将其建模成超图,通过对偶图注意力机制来对以上关联信息进行聚合;最终将BERT和超图对偶注意力网络两个模块提取出的特征进行拼接,经过softmax层得到对文本情感倾向的预测结果。该模型在电商评论二分类数据集和微博文本六分类数据集上的准确率分别达到95.49%和79.83%,相较于基准模型分别提高2.27%~3.45%和6.97%~11.69%;同时还设计了消融实验验证模型各部分对分类结果的增益。实验结果表明,该模型能够显著提高针对中文网络短文本情感分析的准确率。

基于联合GLMB滤波器的可分辨群目标跟踪

针对联合广义标签多伯努利(joint generalized labeled multi-Bernoulli, J-GLMB)滤波算法中群目标之间距离较近、容易关联错误的问题,结合超图匹配(hypergraph matching, HGM)提出一种基于HGM-J-GLMB滤波器的可分辨群目标跟踪算法。首先,采用J-GLMB滤波器估计群内各目标的状态、数目及轨迹信息,并利用HGM结果提升量测与预测状态之间的关联性能。其次,通过图理论计算邻接矩阵,获取群结构信息和子群数目。随后,利用群结构信息估计协作噪声,进而校正目标的预测状态。最后,通过平滑算法改善滤波效果,并设置轨迹长度阈值,使其在平滑状态达到消除短轨迹的目的。仿真实验表明,所提算法在线性系统下能有效提升群目标跟踪性能。

多层次特征融合与超图卷积的生成对抗壁画修复

针对现有壁画深度学习修复方法,存在上下文信息关注不足及结果欠佳的问题,提出了一种多层次特征融合与超图卷积的生成对抗修复模型。首先,利用金字塔特征分层对壁画进行多尺度层次特征提取,并采用混合空洞卷积单元扩大多层特征提取感受野,以克服单尺度卷积操作对于壁画特征提取能力不足的问题。然后,提出多分支短链融合层及门控机制融合多分支特征方法,将相邻分支间的特征信息进行融合,使融合后的壁画特征图中既有同分支的特征,又有相邻分支的特征,以提高特征信息的利用率;并引入门控机制对特征进行选择融合,以减少细节信息的丢失。接着,将融合特征通过卷积长短期记忆网络(ConvLSTM)特征注意力方法,增强对壁画上下文信息的关注。最后,设计超图卷积壁画长程特征增强模块,通过在编码器和解码器的跳跃连接之间建立超图卷积层,利用超图卷积捕获编码器的空间特征信息,并将其迁移到解码器中,有助于解码器更好地生成壁画图像,以加强特征的长程依赖关系,并与SN-PatchGAN判别器对抗博弈从而完成修复。通过敦煌壁画数字化修复实验,结果表明:所提方法客观评价优于对比算法,对于破损壁画修复结果更加清晰自然。

基于超图和MuSig2聚合签名的联盟链主从多链共识机制

针对多链式区块链采用主链最终共识机制,导致主链负载大,制约从链性能等问题,论文提出一种基于超图和MuSig2聚合签名的联盟链主从多链共识机制.首先根据超图理论,构建以横贯超图为主链,子超图为从链的联盟链主从多链架构;然后借鉴分治思想,结合“背书-排序-验证”的共识方式,构建分层分类共识机制,通过分类处理交易降低主链负载压力;最后构建基于MuSig2聚合签名的联盟链多方背书签名方法,提升背书签名的验证效率.性能分析表明:基于MuSig2聚合签名的联盟链多方背书签名安全可靠,基于超图和MuSig2聚合签名的分层分类共识机制具有强一致性和线性时间复杂度.实验结果表明:基于MuSig2聚合签名的多方背书方法的总效率是椭圆曲线数字签名算法(Elliptic Curve Digital Signature Algorithm,ECDSA)的1.55倍,分层分类共识机制能够提升12.5%的共识效率.该机制具有较高性能,可满足企业多样化业务需求.

超图神经网络综述

近年来,图神经网络借助大量数据和超强计算能力在推荐系统和自然语言处理等应用领域取得显著成效,它主要处理具有成对关系的图数据.但许多现实网络中的对象之间的关系是复杂的非成对关系,如科研合作网络、蛋白质网络等.若直接用图结构将这种复杂的关系表示为成对关系,会导致信息丢失.超图是一种灵活的建模工具,可以展现出图无法完整刻画的高阶关系,弥补了图的不足.鉴于此,研究者开始关心如何在超图上设计神经网络,并相继提出应用于下游任务的超图神经网络模型(hypergraph neural network,HGNNs).故对现有的超图神经网络模型进行综述,首先全面回顾超图神经网络在过去3年的研究历程;其次根据设计超图神经网络采用的方法不同对其进行分类,并详细地阐述代表性的模型;然后介绍了超图神经网络的应用领域;最后总结和探讨了超图神经网络未来的研究方向.

联合图随机游走和跳跃连接的动态超图神经网络

针对传统超图神经网络难以提取节点直接邻域外关联度高的节点特征,导致全局特征信息不完整的问题,对动态超图神经网络(DHGNN)进行改进,提出联合图随机游走和跳跃连接的动态超图神经网络(RWS-DHGNN),用于非欧几里得数据的分类。该网络在DHGNN的基础上,引入了图随机游走,从而有效地获取直接邻域外关联度高的节点特征。同时,引入残差网络的思想在超图的顶点卷积处增加跳跃连接构成残差结构。所提网络模型充分发挥图结构和超图结构的优势。在Cora数据集的标准分割和随机分割上将所提网络与GCN、HGNN、GAT和DHGNN进行对比实验,实验结果表明,该模型可以有效提高分类准确率。

面向超图的可解释性对比元路径群组推荐

在群组推荐中庞大且稀疏的数据往往容易忽视用户群组及项目之间的复杂依赖关系,因此融合不同用户偏好行为嵌入,使用户对群组依赖关系的表现更直观,同时为了在对比中增强视图效果,以获得更准确的推荐结果的目的,提出了一个面向超图的可解释性对比元路径群组推荐框架。通过聚合用户项目群组之间的依赖关系,构建元路径表现实体之间的不同类型交互,以促进实体的相似性,更准确地从数据中获取用户的组内、组外交互;通过将可解释性模型与对比学习相结合的技术,以提高模型的可解释性和性能;通过解释引导增强操作在模型框架上生成的正负视图上结合自监督对比学习,来解决上述问题。在真实数据集上进行实验,验证了所提出方法的有效性。