Hypergraph Learning Research Articles

A Review of Hypergraph Neural Networks

In recent years, Graph Neural Networks (GNNs) have seen notable success in fields such as recommendation systems and natural language processing, largely due to the availability of vast amounts of data and powerful computational resources. GNNs are primarily designed to work with graph data that involve pairwise relationships. However, in many real-world networks, the relationships between entities are complex and go beyond simple pairwise connections, as seen in scientific collaboration networks, protein networks, and similar domains. If these complex relationships are directly represented as pairwise relationships using graph structures, it can lead to information loss. A hypergraph, as a special kind of graph-structured data, can represent higher-order relationships that cannot be fully captured by graphs, thereby addressing the limitations of graphs. In light of this, researchers have begun to focus on how to design neural networks on hypergraphs, leading to the proposal of hypergraph neural network (HGNN) models for downstream tasks. Therefore, this paper reviews the existing hypergraph neural network models. The review is conducted from two perspectives: spectral analysis methods and neural network methods on hypergraphs, discussing both unfolded and non-unfolded methods, and further subdividing them based on their algorithm characteristics and application scenarios. Subsequently, the design concepts of various algorithms are analyzed and compared, and the advantages and disadvantages of each type of algorithm are summarized based on experimental results. Finally, potential future research directions in hypergraph learning are discussed.

Distance Enhanced Hypergraph Learning for Dynamic Node Classification

Dynamic node classification aims to predict the labels of nodes in the dynamic networks. Existing methods primarily utilize the graph neural networks to acquire the node features and original graph structure features. However, these approaches ignore the high-order relationships between nodes and may lead to the over-smoothing issue. To address these issues, we propose a distance enhanced hypergraph learning (DEHL) method for dynamic node classification. Specifically, we first propose a time-adaptive pre-training component to generate the time-aware representations of each node. Then we utilize a dual-channel convolution module to construct the local and global hypergraphs which contain the corresponding local and global high-order relationships. Moreover, we adopt the K-nearest neighbor algorithm to construct the global hypergraph in the embedding space. After that, we adopt the node convolution and hyperedge convolution to aggregate the features of neighbors on the hypergraphs to the target node. Finally, we combine the temporal representations and the distance enhanced representations of the target node to predict its label. In addition, we conduct extensive experiments on two public dynamic graph datasets, i.e., Wikipedia and Reddit. The experimental results show that DEHL outperforms the state-of-the-art baselines in terms of AUC.

CHGNN: A Semi-Supervised Contrastive Hypergraph Learning Network

CHGNN: A Semi-Supervised Contrastive Hypergraph Learning Network

The Structure-sharing Hypergraph Reasoning Attention Module for CNNs

Attention mechanisms improve the performance of models by selectively processing relevant information. However, existing attention mechanisms for CNNs do not utilize the high-order semantic similarity between different channels in the input when inferring attention. To address this issue, in this paper, we propose the Structure-sharing Hypergraph Reasoning Attention Module (SHRA Module) to explore the high-order similarity among nodes via hypergraph learning. SHRA Module transforms the input CNN feature maps into hypergraph node representations, which are used to reason attention under a set of learnable hypergraph convolutions. When performing the hypergraph convolution, the SHRA Module utilizes our proposed structure-sharing hypergraph convolution (SHGCN) to perform hypergraph convolutions, where the hypergraphs from different groups and the weight matrices for hypergraph convolutions are conducted in a right-shifted-permutation sequence of hypergraphs. As a result, the weights matrices can be shared with all groups of hypergraphs while performing hypergraph convolution, thus the global information can be used by the module to have a deep look into the input feature. We evaluate SHRA Module with models in object detection, lesion segmentation, and image classification tasks to demonstrate its effectiveness. Experimental results show that SHRA Module highly significantly enhances model performance, surpassing that of classic attention modules.

Multi-view hypergraph regularized Lp norm least squares twin support vector machines for semi-supervised learning

In recent years, multi-view semi-supervised learning has gradually become a popular research direction. The classic binary classification methods in this field are multi-view Laplacian support vector machines (MvLapSVM) and multi-view Laplacian twin support vector machines (MvLapTSVM), which extend semi-supervised support vector machine to multi-view learning. Nevertheless, similar to the majority of SVM-based multi-view methods, the above methods are two-view methods that cannot fully leverage the information from all views and are constructed based on the L2 norm. Additionally, in semi-supervised graph learning, the quality of the graph often has a significant impact on the results. Therefore, we propose a novel multi-view hypergraph regularized Lp norm least squares twin support vector machines (MvHGLpLSTSVM) that can handle general multi-view data for semi-supervised learning. It extends hypergraph learning to multi-view learning and combines Lp norm to further explore the manifold structure and embedded geometric information of multi-view data. By using equality constraints, we design a simple and effective iterative algorithm. In the classification of six multi-view datasets, we compare the proposed method with some other state-of-the-art methods, and the results show that the proposed method is effective.

HGTMDA: A Hypergraph Learning Approach with Improved GCN-Transformer for miRNA-Disease Association Prediction.

Accumulating scientific evidence highlights the pivotal role of miRNA-disease association research in elucidating disease pathogenesis and developing innovative diagnostics. Consequently, accurately identifying disease-associated miRNAs has emerged as a prominent research topic in bioinformatics. Advances in graph neural networks (GNNs) have catalyzed methodological breakthroughs in this field. However, existing methods are often plagued by data noise and struggle to effectively integrate local and global information, which hinders their predictive performance. To address this, we introduce HGTMDA, an innovative hypergraph learning framework that incorporates random walk with restart-based association masking and an enhanced GCN-Transformer model to infer miRNA-disease associations. HGTMDA starts by constructing multiple homogeneous similarity networks. A novel enhancement of our approach is the introduction of a restart-based random walk association masking strategy. By stochastically masking a subset of association data and integrating it with a GCN enhanced by an attention mechanism, this strategy enables better capture of key information, leading to improved information utilization and reduced impact of noisy data. Next, we build an miRNA-disease heterogeneous hypergraph and adopt an improved GCN-Transformer encoder to effectively solve the effective extraction of local and global information. Lastly, we utilize a combined Dice cross-entropy (DCE) loss function to guide the model training and optimize its performance. To evaluate the performance of HGTMDA, comprehensive comparisons were conducted with state-of-the-art methods. Additionally, in-depth case studies on lung cancer and colorectal cancer were performed. The results demonstrate HGTMDA's outstanding performance across various metrics and its exceptional effectiveness in real-world application scenarios, highlighting the advantages and value of this method.

Heterogeneous hypergraph learning for literature retrieval based on citation intents

Heterogeneous hypergraph learning for literature retrieval based on citation intents

Traffic Anomaly Detection based on Spatio-Temporal Hypergraph Convolution Neural Networks

Traffic Anomaly Detection (TAD) is an important and difficult task in Intelligent Transportation Systems (ITS) . Traffic anomaly events are sparse in both spatial and temporal spaces, posing a challenge to the performance of model. Moreover, a single traffic anomaly event can impact multiple road sections in the neighborhood, further undermining the accuracy of TAD. In this paper, we propose a new TAD method based on spatio-temporal hypergraph convolutional neural network. Specifically, we adopt a spatial–temporal augmentation approach for traffic data. This will enhance the performance of detecting sparse anomalies. Meanwhile, we introduce a hypergraph learning method to model the road network. This could capture the spreading features of anomalies for better detection results. Additionally, we design a dynamic hypergraph construction method to extract the evolving relationships of road segments. The proposed model evaluation on the Beijing (SE-BJ) dataset for TAD reveals superior performance compared to state-of-the-art ones.

Integration of protein sequence and protein–protein interaction data by hypergraph learning to identify novel protein complexes

Integration of protein sequence and protein–protein interaction data by hypergraph learning to identify novel protein complexes

Multiple stocks recommendation: a spatio-temporal hypergraph learning approach

Multiple stocks recommendation: a spatio-temporal hypergraph learning approach

MvStHgL: Multi-view Hypergraph Learning with Spatial-temporal Periodic Interests for Next POI Recommendation

Providing potential next point-of-interest (POI) suggestions for users has become a prominent task in location-based social networks, which receives more and more attention from the industry and academia, and it remains challenging due to highly dynamic and personalized interactions in user movements. Currently, state-of-the-art works develop various graph- and sequential-based learning methods to model user-POI interactions and transition regularities. However, there are still two significant shortcomings in these works: (1) Ignoring personalized spatial- and temporal-aspect interactive characteristics capable of exhibiting periodic interests of users; (2) Insufficiently leveraging the sequential patterns of interactions for beyond-pairwise high-order collaborative signals among users’ sequences. To jointly address these challenges, we propose a novel multi-view hypergraph learning with spatial-temporal periodic interests for next POI recommendation (MvStHgL). In the local view, we attempt to learn the POI representation of each interaction via jointing periodic characteristics of spatial and temporal aspects. In the global view, we design a hypergraph by regarding interactive sequences as hyperedges to capture high-order collaborative signals across users, for further POI representations. More specifically, the output of POI representations in the local view is used for the initialized embedding, and the aggregation and propagation in the hypergraph are performed by a novel Node-to-Hypergraph-to-Node scheme. Furthermore, the captured POI embeddings are applied to achieve sequential dependency modeling for next POI prediction. Extensive experiments on three real-world datasets demonstrate that our proposed model outperforms the state-of-the-art models.

FastHGNN: A New Sampling Technique for Learning with Hypergraph Neural Networks

Hypergraphs can represent higher-order relations among objects. Traditional hypergraph neural networks involve node-edge-node transform, leading to high computational cost and timing. The main aim of this paper is to propose a new sampling technique for learning with hypergraph neural networks. The core idea is to design a layer-wise sampling scheme for nodes and hyperedges to approximate the original hypergraph convolution. We rewrite hypergraph convolution in the form of double integral and leverage Monte Carlo to achieve a discrete and consistent estimator. In addition, we use importance sampling and finally derive feasible probability mass functions for both nodes and hyperedges in consideration of variance reduction, based on some assumptions. Notably, the proposed sampling technique allows us to handle large-scale hypergraph learning, which is not feasible with traditional hypergraph neural networks. Experiment results demonstrate that our proposed model keeps a good balance between running time and prediction accuracy.

Analyzing the packaging design evaluation based on image emotion perception computing

Nowadays, a wide variety of labels of items are widely available, and human consumption is increasingly tailored to meet their individual needs. So, many businesses are starting to focus on improving the functionality of modern packaging. Sensorial paradigms and emotional reactions could change during the user-product interaction lifecycle. The designer's emotional imagination and past experiences are the backbone of conventional product package design, which has limitations due to unmanageable content and an absence of professional advice—the majority of previous research on emotional image analysis aimed to forecast the most common viewer emotions. Since the feelings a picture evokes are quite individual and vary from viewer to viewer, this overarching feeling isn't always enough for practical uses. The research presented an approach to packaging design evaluation based on image emotion perception computing (PDE-IEPC), which combines emotion perception technology with a deep LSTM (Long short-term model), resulting in an immersive and dynamic experience for the human senses. Emotion Perception Computing's Dynamic Multi-task Hypergraph Learning (DMHL) approach considers graphical data, social context, spatial evolution, and location, among other criteria, to evaluate packaging designs efficiently based on their emotional impact. Image-Emotion-Social-Net is a large dataset used to evaluate multidimensional and categorical attitude representation. The dataset is sourced from Flickr and contains over 1 million images presented by over 9000 users. Personalized emotion categorization is an area where research on this dataset shows that the suggested strategy outperforms many modern techniques. The experimental results show that the proposed method achieves a high packaging design quality rate of 94.1 %, a performance success rate of 97.5 %, and a mean square error rate of 2 % compared to other existing methods.

Exploring High-Order Skeleton Correlations with Physical and Non-Physical Connection for Action Recognition

Hypergraphs have received widespread attention in modeling complex data correlations due to their superior performance. In recent years, some researchers have used hypergraph structures to characterize complex non-pairwise joints in the human skeleton and model higher-order correlations of the human skeleton. However, traditional methods of constructing hypergraphs based on physical connections ignore the dependencies among non-physically connected joints or bones, and it is difficult to model the correlation among joints or bones that are highly correlated in human action but are physically connected at long distances. To address these issues, we propose a skeleton-based action recognition method for hypergraph learning based on skeleton correlation, which explores the effects of physically and non-physically connected skeleton information on accurate action recognition. Specifically, in this paper, spatio-temporal correlation modeling is performed on the natural connections inherent in humans (physical connections) and the joints or bones that are more dependent but not directly connected (non-physical connection) during human actions. In order to better learn the hypergraph structure, we construct a spatio-temporal hypergraph neural network to extract the higher-order correlations of the human skeleton. In addition, we use an attentional mechanism to compute the attentional weights among different hypergraph features, and adaptively fuse the rich feature information in different hypergraphs. Extensive experiments are conducted on two datasets, NTU-RGB+D 60 and Kinetics-Skeleton, and the results show that compared with the state-of-the-art skeleton-based methods, our proposed method can achieve an optimal level of performance with significant advantages, providing a more accurate environmental perception and action analysis for the development of embodied intelligence.

DHyper: A Recurrent Dual Hypergraph Neural Network for Event Prediction in Temporal Knowledge Graphs

Event prediction is a vital and challenging task in temporal knowledge graphs (TKGs), which have played crucial roles in various applications. Recently, many graph neural networks based approaches are proposed to model the graph structure information in TKGs. However, these approaches only construct graphs based on quadruplets and model the pairwise correlation between entities, which fail to capture the high-order correlations among entities. To this end, we propose DHyper, a recurrent Dual Hypergraph neural network for event prediction in TKGs, which simultaneously models the influences of the high-order correlations among both entities and relations. Specifically, a dual hypergraph learning module is proposed to discover the high-order correlations among entities and among relations in a parameterized way. A dual hypergraph message passing network is introduced to perform the information aggregation and representation fusion on the entity hypergraph and the relation hypergraph. Extensive experiments on six real-world datasets demonstrate that DHyper achieves the state-of-the-art performances, outperforming the best baseline by an average of 13.09%, 4.26%, 17.60%, and 18.03% in MRR, Hits@1, Hits@3, and Hits@10, respectively.

Hypergraph regularized nonnegative triple decomposition for multiway data analysis

Tucker decomposition is widely used for image representation, data reconstruction, and machine learning tasks, but the calculation cost for updating the Tucker core is high. Bilevel form of triple decomposition (TriD) overcomes this issue by decomposing the Tucker core into three low-dimensional third-order factor tensors and plays an important role in the dimension reduction of data representation. TriD, on the other hand, is incapable of precisely encoding similarity relationships for tensor data with a complex manifold structure. To address this shortcoming, we take advantage of hypergraph learning and propose a novel hypergraph regularized nonnegative triple decomposition for multiway data analysis that employs the hypergraph to model the complex relationships among the raw data. Furthermore, we develop a multiplicative update algorithm to solve our optimization problem and theoretically prove its convergence. Finally, we perform extensive numerical tests on six real-world datasets, and the results show that our proposed algorithm outperforms some state-of-the-art methods.

EduCross: Dual adversarial bipartite hypergraph learning for cross-modal retrieval in multimodal educational slides

In the digital education landscape, cross-modal retrieval (CMR) from multimodal educational slides represents a significant challenge, particularly because of the complex nature of academic content, which includes images, diagrams, equations, and tables across various subjects such as mathematics and biology. Current CMR systems are primarily designed for “(natural) image to text” interactions (or vice versa) and inadequately address real-world educational scenarios. This study presents EduCross, a novel framework devised to enhance CMR within multimodal educational slides, which is a domain in which traditional retrieval systems fall short. Recognizing the imperative for a system that is tailored to the educational context, EduCross integrates dual adversarial bipartite hypergraph learning, harnessing the capabilities of generative adversarial networks with figure-text dual channels. This powerful combination facilitates robust bidirectional mapping, allowing for the precise association of figures with their descriptive spoken language segments and ensuring a comprehensive CMR experience. Specifically, we develop framelet-based deep bipartite hypergraph neural networks that effectively manage the high-order relationships between diverse educational content types and various types of slide figures. Our experimental results underscore the superior performance of EduCross, demonstrating its effectiveness through the use of the real Multimodal Lecture Presentations dataset that mirrors authentic educational settings. These outcomes highlight the significant advancements of EduCross over existing methods, marking a leap forward in the accurate retrieval of multimodal educational content.

Advancing urban traffic accident forecasting through sparse spatio-temporal dynamic learning

Traffic accidents have emerged as one of the most public health safety matters, raising concerns from both the public and urban administrators. The ability to accurately predict traffic accident not only supports the governmental decision-making in advance but also enhances public confidence in safety measures. However, the efficacy of traditional spatio-temporal prediction models are compromised by the skewed distributions and sparse labeling of accident data. To this end, we propose a Sparse Spatio-Temporal Dynamic Hypergraph Learning (SST-DHL) framework that captures higher-order dependencies in sparse traffic accidents by combining hypergraph learning and self-supervised learning. The SST-DHL model incorporates a multi-view spatiotemporal convolution block to capture local correlations and semantics of traffic accidents, a cross-regional dynamic hypergraph learning model to identify global spatiotemporal dependencies, and a two-supervised self-learning paradigm to capture both local and global spatiotemporal patterns. Through experimentation on New York City and London accident datasets, we demonstrate that our proposed SST-DHL exhibits significant improvements compared to optimal baseline models at different sparsity levels. Additionally, it offers enhanced interpretability of results by elucidating complex spatio-temporal dependencies among various traffic accident instances. Our study demonstrates the effectiveness of the SST-DHL framework in accurately predicting traffic accidents, thereby enhancing public safety and trust.

Self-Supervised Hypergraph Learning for Knowledge-Aware Social Recommendation

Social recommendations typically utilize social relationships and past behaviors to predict users’ preferences. In real-world scenarios, the connections between users and items often extend beyond simple pairwise relationships. Leveraging hypergraphs to capture high-order relationships provides a novel perspective to social recommendation. However, effectively modeling these high-order relationships is challenging due to limited external knowledge and noisy feedback. To tackle these challenges, we propose a novel framework called self-supervised hypergraph learning for knowledge-aware social recommendation (SHLKR). In SHLKR, we incorporated three main types of connections: behavior, social, and attribute context relationships. These dependencies serve as the basis for defining hyperedges in the hypergraphs. A dual-channel hypergraph structure is created based on these relationships. Then, the hypergraph convolution is applied to model the high-order interactions between users and items. Additionally, we adopted a self-supervised learning task to maximize the consistency between different views. It helps to mitigate the model’s sensitivity to noisy feedback. We evaluated the performance of SHLKR through extensive experiments on publicly available datasets. The results demonstrate that leveraging hypergraphs for modeling can better capture the complexity and diversity of user preferences and interactions.

THGFormer: Time-Aware Hypergraph Learning for Multimodal Social Media Popularity Prediction (Student Abstract)

Social media popularity prediction of multimodal user-generated content (UGC) is a crucial task for many real-world applications. However, existing efforts are often limited by missing inter-instance correlations and UGC temporal patterns. To address these issues, we propose a novel time-aware hypergraph Transformer framework, THGFormer. It fully represents inter-instance and intra-instance relations by hypergraphs, captures the temporal dependencies with a time encoder, and enhances UGC's representations via a neighborhood knowledge aggregation. Extensive experiments conducted on two real-world datasets demonstrate that THGFormer outperforms state-of-the-art popularity prediction models across several settings.