Global Structural Features Research Articles

Citation and bibliographic coupling between authors in the field of social network analysis

Abstract Purpose We analyzed the structure of a community of authors working in the field of social network analysis (SNA) based on citation indicators: direct citation and bibliographic coupling metrics. We observed patterns at the micro, meso, and macro levels of analysis. Design/methodology/approach We used bibliometric network analysis, including the “temporal quantities” approach proposed to study temporal networks. Using a two-mode network linking publications with authors and a one-mode network of citations between the works, we constructed and analyzed the networks of citation and bibliographic coupling among authors. We used an iterated saturation data collection approach. Findings At the macro-level, we observed the global structural features of citations between authors, showing that 80% of authors have not more than 15 citations from other works. At the meso-level, we extracted the groups of authors citing each other and similar to each other according to their citation patterns. We have seen a division of authors in SNA into groups of social scientists and physicists, as well as into other groups of authors from different disciplines. We found some examples of brokerage between different groups that maintained the common identity of the field. At the micro-level, we extracted authors with extremely high values of received citations, who can be considered as the most prominent authors in the field. We examined the temporal properties of the most popular authors. Research limitations The main challenge in this approach is the resolution of the author’s name (synonyms and homonyms). We faced the author disambiguation, or “multiple personalities” (Harzing, 2015) problem. To remain consistent and comparable with our previously published articles, we used the same SNA data collected up to 2018. The analysis and conclusions on the activity, productivity, and visibility of the authors are relative only to the field of SNA. Practical implications The proposed approach can be utilized for similar objectives and identifying key structures and characteristics in other disciplines. This may potentially inspire the application of network approaches in other research areas, creating more authors collaborating in the field of SNA. Originality/value We identified and applied an innovative approach and methods to study the structure of scientific communities, which allowed us to get the findings going beyond those obtained with other methods. We used a new approach to temporal network analysis, which is an important addition to the analysis as it provides detailed information on different measures for the authors and pairs of authors over time.

Hierarchical Fusion of Infrared and Visible Images Based on Channel Attention Mechanism and Generative Adversarial Networks.

In order to solve the problem that existing visible and infrared image fusion methods rely only on the original local or global information representation, which has the problem of edge blurring and non-protrusion of salient targets, this paper proposes a layered fusion method based on channel attention mechanism and improved Generative Adversarial Network (HFCA_GAN). Firstly, the infrared image and visible image are decomposed into a base layer and fine layer, respectively, by a guiding filter. Secondly, the visible light base layer is fused with the infrared image base layer by histogram mapping enhancement to improve the contour effect. Thirdly, the improved GAN algorithm is used to fuse the infrared and visible image refinement layer, and the depth transferable module and guided fusion network are added to enrich the detailed information of the fused image. Finally, the multilayer convolutional fusion network with channel attention mechanism is used to correlate the local information of the layered fusion image, and the final fusion image containing contour gradient information and useful details is obtained. TNO and RoadSence datasets are selected for training and testing. The results show that the proposed algorithm retains the global structure features of multilayer images and has obvious advantages in fusion performance, model generalization and computational efficiency.

Global and local feature extraction based on convolutional neural network residual learning for MR image denoising.

Given the different noise distribution information of global and local magnetic resonance (MR) images, this study aims to extend the current work on convolutional neural networks that preserve global structure and local details in MR image denoising tasks. This study proposed a parallel and serial network for denoising 3D MR images, called 3D-PSNet. We use the residual depthwise separable convolution block to learn the local information of the feature map, reduce the network parameters, and thus improve the training speed and parameter efficiency. In addition, we consider the feature extraction of the global image and utilize residual dilated convolution to process the feature map to expand the receptive field of the network and avoid the loss of global information. Finally, we combine both of them to form a parallel network. What's more, we integrate reinforced residual convolution blocks with dense connections to form serial network branches, which can remove redundant information and refine features to further obtain accurate noise information. The peak signal-to-noise ratio, structural similarity index measure, and root mean square error metrics of 3D-PSNet are as high as 47.79%, 99.81%, and 0.40%, respectively, achieving competitive denoising effect on three public datasets. The ablation experiments demonstrated the effectiveness of all the designed modules regarding all the evaluated metrics in both datasets. The proposed 3D-PSNet takes advantage of multi-scale receptive fields, local feature extraction and residual dense connections to more effectively restore the global structure and local fine features in MR images, and is expected to help doctors quickly and accurately diagnose patients' conditions.

Probing microstructure evolution of Si/C anode for Li-ion batteries via synchrotron transmission X-ray tomographic microscopy

Holding great potentials as an excellent anode for Li-ion batteries, Si has received enormous attention, and its degradation theory has been enriched extensively based on various characterizations over time. To provide a more comprehensive diagnosis for current Si/C anodes, ex situ phase-contrast synchrotron transmission X-ray tomographic microscopy (TXTM) is applied to study a Si/C electrode sampled at representative cycling numbers. Through a series of data segmentations and identifications, three-dimensional Si/C electrodes are described in details with both global and localized structure features, which are rationally linked to the electrochemical characteristics as well as the previous discoveries. This successfully establishes an updated electrode-level failure process for common Si/C anodes where the electrode degradation is primarily initiated by the nonstop Si-phase pulverization with continually-expanded LixSi/electrolyte interface, and manifested as the internal void space expansion and the ever-growing electrode thickness. Bearing this in mind, multiple complementary and synergistic solutions towards limiting the intrinsic Si rupture and electrolyte decomposition should be developed to ensure a high-performance Si-based anode over long-term cycling for Li-ion battery in the near future.

ZOLLINGER LAMELLA VAULT OF EXISTING TIMBER HANGARS IN ŠIBENIK (CASE STUDY) – INITIAL CONDITION ASSESSMENT AND PRELIMINARY ANALYSIS OF GLOBAL STRUCTURAL FEATURES

<p class="KEYWORDS" style="text-align: justify;">A preliminary condition assessment using non-destructive testing methods for determining material quality, has been carried out on three adjacent wooden hangars near Šibenik with the aim of exploring the possibilities of their repurposing. Through visual inspection and detailed geometric measurements, it has been verified that their unique structure embodies the original Zollinger lamella vault system in which the joint typology is decisive for both, the overarching structural analysis and the execution of the structure. In this paper, the results of conducted investigation, including the developed detailed geometric model are presented, and guidelines for future comprehensive research focusing on the influence of joint stiffness on global structural features are discussed, as well.</p>

Multi-Threshold Recurrence Rate Plot: A Novel Methodology for EEG Analysis in Alzheimer's Disease and Frontotemporal Dementia.

This study introduces Multi-Threshold Recurrence Rate Plots (MTRRP), a novel methodology for analyzing dynamic patterns in complex systems, such as those influenced by neurodegenerative diseases in brain activity. MTRRP characterizes how recurrence rates evolve with increasing recurrence thresholds. A key innovation of our approach, Recurrence Complexity, captures structural complexity by integrating local randomness and global structural features through the product of Recurrence Rate Gradient and Recurrence Hurst, both derived from MTRRP. We applied this technique to resting-state EEG data from patients diagnosed with Alzheimer's Disease (AD), Frontotemporal Dementia (FTD), and age-matched healthy controls. The results revealed significantly higher recurrence complexity in the occipital areas of AD and FTD patients, particularly pronounced in the Alpha and Beta frequency bands. Furthermore, EEG features derived from MTRRP were evaluated using a Support Vector Machine with leave-one-out cross-validation, achieving a classification accuracy of 87.7%. These findings not only underscore the utility of MTRRP in detecting distinct neurophysiological patterns associated with neurodegenerative diseases but also highlight its broader applicability in time series analysis, providing a substantial tool for advancing medical diagnostics and research.

DSSGNN-PPI: A Protein–Protein Interactions prediction model based on Double Structure and Sequence graph neural networks

The process of experimentally confirming complex interaction networks among proteins is time-consuming and laborious. This study aims to address Protein–Protein Interactions (PPIs) prediction based on graph neural networks (GNN). A novel multilevel prediction model for PPIs named DSSGNN-PPI (Double Structure and Sequence GNN for PPIs) is designed. Initially, a distance graph between amino acid residues is constructed. Subsequently, the distance graph is fed into an underlying graph attention network module. This enables us to efficiently learn vector representations that encode the three-dimensional structure of proteins and simultaneously aggregate key local patterns and overall topological information to obtain graph embedding that adequately represent local and global structural features. In addition, the embedding representations that reflect sequence properties are obtained. Two features are fused to construct high-level protein complex networks, which are fed into the designed gated graph attention network to extract complex topological patterns. By combining heterogeneous multi-source information from downstream structure graph and upstream sequence models, the understanding of PPIs is comprehensively enhanced. A series of evaluation results validate the remarkable effectiveness of DSSGNN-PPI framework in enhancing the prediction of multi-type interactions among proteins. The multilevel representation learning and information fusion strategies provide a new effective solution paradigm for structural biology problems. The source code for DSSGNN-PPI has been hosted on GitHub and is available at https://github.com/cstudy1/DSSGNN-PPI.

Reconstruction from CT truncated data based on dual-domain transformer coupled feature learning

To propose a CT truncated data reconstruction model (DDTrans) based on projection and image dualdomain Transformer coupled feature learning for reducing truncation artifacts and image structure distortion caused by insufficient field of view (FOV) in CT scanning. Transformer was adopted to build projection domain and image domain restoration models, and the long-range dependency modeling capability of the Transformer attention module was used to capture global structural features to restore the projection data information and enhance the reconstructed images. We constructed a differentiable Radon back-projection operator layer between the projection domain and image domain networks to enable end-to-end training of DDTrans. Projection consistency loss was introduced to constrain the image forwardprojection results to further improve the accuracy of image reconstruction. The experimental results with Mayo simulation data showed that for both partial truncation and interior scanning data, the proposed DDTrans method showed better performance than the comparison algorithms in removing truncation artifacts at the edges and restoring the external information of the FOV. The DDTrans method can effectively remove CT truncation artifacts to ensure accurate reconstruction of the data within the FOV and achieve approximate reconstruction of data outside the FOV.

The Attention-Based Autoencoder for Network Traffic Classification with Interpretable Feature Representation

Network traffic classification is crucial for identifying network applications and defending against network threats. Traditional traffic classification approaches struggle to extract structural features and suffer from poor interpretability of feature representations. The high symmetry between network traffic classification and its interpretable feature representation is vital for network traffic analysis. To address these issues, this paper proposes a traffic classification and feature representation model named the attention mechanism autoencoder (AMAE). The AMAE model extracts the global spatial structural features of network traffic through attention mechanisms and employs an autoencoder to extract local structural features and perform dimensionality reduction. This process maps different network traffic features into one-dimensional coordinate systems in the form of spectra, termed FlowSpectrum. The spectra of different network traffic represent different intervals in the coordinate system. This paper tests the interpretability and classification performance of network traffic features of the AMAE model using the ISCX-VPN2016 dataset. Experimental results demonstrate that by analyzing the overall distribution of attention weights and local weight values of network traffic, the model effectively explains the differences in the spectral representation intervals of different types of network traffic. Furthermore, our approach achieves the highest classification accuracy of up to 100% for non-VPN-encrypted traffic and 99.69% for VPN-encrypted traffic, surpassing existing traffic classification schemes.

Fusion of satellite and street view data for urban traffic accident hotspot identification

As the number of vehicles and the volume of traffic swell in urban centers, cities have experienced a concomitant increase in traffic accidents. Proactively identifying accident-prone hotspots in urban environments holds the promise of preventing traffic mishaps, thereby curtailing the incidence of accidents and reducing property damage. This research introduces the Two-Branch Contextual Feature-Guided Converged Network (TCFGC-Net) utilizing multimodal satellite and street view data. Designed to extract global structural features from satellite imagery and dynamic continuous features from street view imagery, the model aims to improve the accuracy of detecting urban accident hotspots. For the satellite imagery branch, we propose the Contextual Feature Coupled Convolutional Neural Network (Trans-CFCCNN) designed to extract global spatial features and discern feature correlations across adjacent regions. For the street view imagery branch, we develop the Sequential Feature Recurrent Attention Network (SFRAN) to assimilate and integrate dynamic scene features captured from successive street view images. We designed the Multi-Branch Feature Adaptive Fusion Structure (MBFAF) to aggregate different branch features for accurate identification of accident hotspots. Experimental results show that the model performs well, with an overall accuracy of 93.7 %. Ablation studies confirm that relative to standalone street view and satellite branch analyses, implementing multimodal fusion enhances the model's accuracy by 12.05 % and 17.86 %, respectively. The innovative fusion structure proposed herein garners a 4.22 % increase in model accuracy, outpacing conventional feature concatenation techniques. Furthermore, the model outperforms existing deep learning models in terms of overall efficacy. Additionally, to showcase the efficacy of the proposed model structure, we utilize Class Activation Maps (CAM) to provide visual interpretability for the model. These results suggest that the dual-branch fusion model effectively decreases false alarm occurrences and directs the model's focus toward regions more pertinent to accident hotspots. Finally, the code and model used for identifying hotspots of urban traffic accidents in this study are available for access: https://github.com/gwt-ZJU/TCFGC-Net.

Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-Based Similarity

Graph representation learning (GRL) methods, such as graph neural networks and graph transformer models, have been successfully used to analyze graph-structured data, mainly focusing on node classification and link prediction tasks. However, the existing studies mostly only consider local connectivity while ignoring long-range connectivity and the roles of nodes. In this paper, we propose Unified Graph Transformer Networks (UGT) that effectively integrate local and global structural information into fixed-length vector representations. First, UGT learns local structure by identifying the local sub-structures and aggregating features of the k-hop neighborhoods of each node. Second, we construct virtual edges, bridging distant nodes with structural similarity to capture the long-range dependencies. Third, UGT learns unified representations through self-attention, encoding structural distance and p-step transition probability between node pairs. Furthermore, we propose a self-supervised learning task that effectively learns transition probability to fuse local and global structural features, which could then be transferred to other downstream tasks. Experimental results on real-world benchmark datasets over various downstream tasks showed that UGT significantly outperformed baselines that consist of state-of-the-art models. In addition, UGT reaches the third-order Weisfeiler-Lehman power to distinguish non-isomorphic graph pairs.

Exploring an accurate machine learning model to quickly estimate stability of diverse energetic materials

High energy and low sensitivity have been the focus of developing new energetic materials (EMs). However, there has been a lack of a quick and accurate method for evaluating the stability of diverse EMs. Here, we develop a machine learning prediction model with high accuracy for bond dissociation energy (BDE) of EMs. A reliable and representative BDE dataset of EMs is constructed by collecting 778 experimental energetic compounds and quantum mechanics calculation. To sufficiently characterize the BDE of EMs, a hybrid feature representation is proposed by coupling the local target bond into the global structure characteristics. To alleviate the limitation of the low dataset, pairwise difference regression is utilized as a data augmentation with the advantage of reducing systematic errors and improving diversity. Benefiting from these improvements, the XGBoost model achieves the best prediction accuracy with R2 of 0.98 and MAE of 8.8kJ mol-1, significantly outperforming competitive models.

Multi-channel high-order network representation learning research.

The existing network representation learning algorithms mainly model the relationship between network nodes based on the structural features of the network, or use text features, hierarchical features and other external attributes to realize the network joint representation learning. Capturing global features of the network allows the obtained node vectors to retain more comprehensive feature information during training, thereby enhancing the quality of embeddings. In order to preserve the global structural features of the network in the training results, we employed a multi-channel learning approach to perform high-order feature modeling on the network. We proposed a novel algorithm for multi-channel high-order network representation learning, referred to as the Multi-Channel High-Order Network Representation (MHNR) algorithm. This algorithm initially constructs high-order network features from the original network structure, thereby transforming the single-channel network representation learning process into a multi-channel high-order network representation learning process. Then, for each single-channel network representation learning process, the novel graph assimilation mechanism is introduced in the algorithm, so as to realize the high-order network structure modeling mechanism in the single-channel network representation learning. Finally, the algorithm integrates the multi-channel and single-channel mechanism of high-order network structure joint modeling, realizing the efficient use of network structure features and sufficient modeling. Experimental results show that the node classification performance of the proposed MHNR algorithm reaches a good order on Citeseer, Cora, and DBLP data, and its node classification performance is better than that of the comparison algorithm used in this paper. In addition, when the vector length is optimized, the average classification accuracy of nodes of the proposed algorithm is up to 12.24% higher than that of the DeepWalk algorithm. Therefore, the node classification performance of the proposed algorithm can reach the current optimal order only based on the structural features of the network under the condition of no external feature supplementary modeling.

Smart Contract Vulnerability Detection Based on Multi-Scale Encoders

Vulnerabilities in smart contracts may trigger serious security events, and the detection of smart contract vulnerabilities has become a significant problem. In this paper, to solve the limitations of current deep learning-based vulnerability detection methods in extracting various code critical features, using the multi-scale cascade encoder architecture as the backbone, we propose a novel Multi-Scale Encoder Vulnerability Detection (MEVD) approach to hit well-known high-risk vulnerabilities in smart contracts. Firstly, we use the gating mechanism to design a unique Surface Feature Encoder (SFE) to enrich the semantic information of code features. Then, by combining a Base Transformer Encoder (BTE) and a Detail CNN Encoder (DCE), we introduce a dual-branch encoder to capture the global structure and local detail features of the smart contract code, respectively. Finally, to focus the model’s attention on vulnerability-related characteristics, we employ the Deep Residual Shrinkage Network (DRSN). Experimental results on three types of high-risk vulnerability datasets demonstrate performance compared to state-of-the-art methods, and our method achieves an average detection accuracy of 90%.

Research on Medical Image Segmentation Method Based on Improved U-Net3.

Computer assisted diagnostic technology has been widely used in clinical practice, specifically focusing on medical image segmentation. Its purpose is to segment targets with certain special meanings in medical images and extract relevant features, providing reliable basis for subsequent clinical diagnosis and research. However, because of different shapes and complex structures of segmentation targets in different medical images, some imaging techniques have similar characteristics, such as intensity, color, or texture, for imaging different organs and tissues. The localization and segmentation of targets in medical images remains an urgent technical challenge to be solved. As such, an improved full scale skip connection network structure for the CT liver image segmentation task is proposed. This structure includes a biomimetic attention module between the shallow encoder and the deep decoder, and the feature fusion proportion coefficient between the two is learned to enhance the attention of the overall network to the segmented target area. In addition, based on the traditional point sampling mechanism, an improved point sampling strategy is proposed for characterizing medical images to further enhance the edge segmentation effect of CT liver targets. The experimental results on the commonly used combined (CT-MR) health absolute organ segmentation (CHAOS) dataset show that the average dice similarity coefficient (DSC) can reach 0.9467, the average intersection over union (IOU) can reach 0.9623, and the average F1 score can reach 0.9351. This indicates that the model can effectively learn image detail features and global structural features, leading to improved segmentation of liver images.

RGCNPPIS: A Residual Graph Convolutional Network for Protein-Protein Interaction Site Prediction.

Accurate identification of protein-protein interaction (PPI) sites is crucial for understanding the mechanisms of biological processes, developing PPI networks, and detecting protein functions. Currently, most computational methods primarily concentrate on sequence context features and rarely consider the spatial neighborhood features. To address this limitation, we propose a novel residual graph convolutional network for structure-based PPI site prediction (RGCNPPIS). Specifically, we use a GCN module to extract the global structural features from all spatial neighborhoods, and utilize the GraphSage module to extract local structural features from local spatial neighborhoods. To the best of our knowledge, this is the first work utilizing local structural features for PPI site prediction. We also propose an enhanced residual graph connection to combine the initial node representation, local structural features, and the previous GCN layer's node representation, which enables information transfer between layers and alleviates the over-smoothing problem. Evaluation results demonstrate that RGCNPPIS outperforms state-of-the-art methods on three independent test sets. In addition, the results of ablation experiments and case studies confirm that RGCNPPIS is an effective tool for PPI site prediction.

Mutations at proximal cysteine residues in PML impair ATO binding by destabilizing the RBCC domain.

Acute promyelocytic leukemia (APL) is characterized by the fusion gene promyelocytic leukemia-retinoic acid receptor-alpha (PML-RARA) and is conventionally treated with arsenic trioxide (ATO). ATO binds directly to the RING finger, B-box, coiled-coil (RBCC) domain of PML and initiates degradation of the fusion oncoprotein PML-RARA. However, the mutational hotspot at C212-S220 disrupts ATO binding, leading to drug resistance in APL. Therefore, structural consequences of these point mutations in PML that remain uncertain require comprehensive analysis. In this study, we investigated the structure-based ensemble properties of the promyelocytic leukemia-RING-B-box-coiled-coil (PML-RBCC) domains and ATO-resistant mutations. Oligomeric studies reveal that PML-RBCC wild-type and mutants C212R, S214L, A216T, L217F, and S220G predominantly form tetramers, whereas mutants C213R, A216V, L218P, and D219H tend to form dimers. The stability of the dimeric mutants was lower, exhibiting a melting temperature (Tm) reduction of 30 °C compared with the tetrameric mutants and wild-type PML protein. Furthermore, the exposed surface of the C213R mutation rendered it more prone to protease digestion than that of the C212R mutation. The spectroscopic analysis highlighted ATO-induced structural alterations in S214L, A216V, and D219H mutants, in contrast to C213R, L217F, and L218P mutations. Moreover, the computational analysis revealed that the ATO-resistant mutations C213R, A216V, L217F, and L218P caused changes in the size, shape, and flexibility of the PML-RBCC wild-type protein. The mutations C213R, A216V, L217F, and L218P destabilize the wild-type protein structure due to the adaptation of distinct conformational changes. In addition, these mutations disrupt several hydrogen bonds, including interactions involving C212, C213, and C215, which are essential for ATO binding. The local and global structural features induced by these mutations provide mechanistic insight into ATO resistance and APL pathogenesis.

SECANet: A structure‐enhanced attention network with dual‐domain contrastive learning for scene text image super‐resolution

AbstractIn this letter, we developed novel Structure Enhanced Channel Attention Network (SECANet) for scene text image super‐resolution (STISR). The newly proposed SECANet integrates a group of Structure‐Enhanced Attention Modules to focus more on both local and global structural features in the character regions of text images. Moreover, we elaborately formulate a Dual‐Domain Contrastive Learning framework that integrates one pixel‐level contrastive loss and the other semantic‐level contrastive loss to jointly optimize the SECANet for generating more visually pleasing yet better recognizable high‐quality SR images without introducing any additional prior generators in both the training and testing stages, showing promising computational efficiency. Experimental results on the Textzoom dataset indicate that our method can achieve both decent performance in super‐resolving more impressive scene text images from low‐resolution ones and better recognition accuracy than other competitors.

Incorporating medical domain knowledge into data-driven method: A vessel attention guided multi-granularity network for automatic cataract classification

Early cataract screening is essential to help patients prevent irreversible vision loss. Existing data-driven methods can automatically extract deep fundus features for automatic cataract classification. However, due to the lack of medical domain knowledge, these deep features may contain large pathological irrelevant information, thus reducing the classification accuracy of data-driven methods. To fully take the advantages of domain knowledge to enhance the discriminative representation and avoid the irrelevant impact, we comprehensively investigate how to incorporate it into data-driven models. Considering the vessel information and multi-granularity representation are two key points of domain knowledge for cataract diagnosis, we propose a vessel attention guided multi-granularity network (VAM-Net), which contains three main components. Specifically, the vessel-level segmentation subnet is firstly designed by leveraging U-shape segmentation network. The refined vessel information obtained in this step can be regarded as visual attention to enable deep network extract discriminative features. With the guidance of vessel attention, the global-local classification subnet is further proposed to eliminate the interference of irrelevant information and enhance the multi-granularity feature learning (including global structural features and local subtle features). Finally, the multiple levels predictions are integrated into the final cataract diagnosis decision via constructing multi-granularity ensemble network. In this manner, both domain knowledge and deep networks are organically combined for pursuing better diagnosis performance. By setting extensive comparative and ablation experiments, we evaluate the effectiveness of the proposed VAM-Net on the real-world cataract dataset (92.78% detection accuracy and 87.68% grading accuracy).

Image emotion multi-label classification based on multi-graph learning

Images contain rich information and can induce various emotions in the audience. Image emotion classification aims to identify the emotion categories that images can evoke. It is widely used in mental health assessment, human–computer interaction, etc. There are two main problems in existing image emotion classification methods: (1) Most of them only focus on a single emotion label; (2) The global structural relationship among semantic objects in the image is ignored. Therefore, this paper proposes an Image Emotion classification method based on Multi-Graph Multi-Label learning (IE-MGML). In contrast to the existing approaches, the image is transformed into a graph-based representation by extracting the emotional features of semantic objects and calculating the similarity between the features. The local (semantic objects) features and global structure (relationship among semantic objects) features of the image are fused by the relationship between nodes. Furthermore, the graph representation of an image from the perspective of multiple emotional features is pooled and modeled as a graph bag containing multiple graphs (i.e., multi-graph). In multi-graph learning, the graph kernel directly evaluates a graph-label dependency score to avoid the loss of structural information caused by graph-instance degradation. The bag(image)-label dependency score is obtained by aggregating the graph-label dependency score from different perspectives through the aggregation function. The problem of error accumulation in the learning process is handled by proposing a threshold-based ranking loss objective function. Moreover, the non-convex optimization problem is addressed using a subgradient descent algorithm to deal with the required high-dimensional space computation. Experimental results on three general image emotion datasets show that the proposed method outperforms the state-of-the-art methods.