The Graph Neural Network Model
Open accessThe DOI matches its Crossref record; 109 of the 111 references it cites were found in a registry or an index.
Abstract
Many underlying relationships among data in several areas of science and engineering, e.g., computer vision, molecular chemistry, molecular biology, pattern recognition, and data mining, can be represented in terms of graphs. In this paper, we propose a new neural network model, called graph neural network (GNN) model, that extends existing neural network methods for processing the data represented in graph domains. This GNN model, which can directly process most of the practically useful types of graphs, e.g., acyclic, cyclic, directed, and undirected, implements a function tau(G,n) is an element of IR(m) that maps a graph G and one of its nodes n into an m-dimensional Euclidean space. A supervised learning algorithm is derived to estimate the parameters of the proposed GNN model. The computational cost of the proposed algorithm is also considered. Some experimental results are shown to validate the proposed learning algorithm, and to demonstrate its generalization capabilities.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“most rely on homogeneous and static structures, which limits their ability to capture the heterogeneity and temporal evolution of real-world environments.”
From A Survey of Heterogeneous Graph Neural Networks for Cybersecurity Anomaly Detection (2025) · cites this paper as [3]
In the citing paper
“Despite notable progress in heterogeneous graph neural networks (HGNNs) for anomaly detection, several obstacles remain before such methods can be reliably deployed in real-world cybersecurity systems [3], [18], [75].”
From A Survey of Heterogeneous Graph Neural Networks for Cybersecurity Anomaly Detection (2025) · cites this paper as [3]
In the citing paper
“After convolution, feature maps consist of a large number of features that is prone to causing overfitting problem [21].”
From A Survey of Convolutional Neural Networks: Analysis, Applications, and Prospects (2022) · cites this paper as [21]
In the citing paper
“Graph neural networks (GNNs) are a type of deep learning model that are specifically designed and well-suited for analyzing complex relationships between objects represented in a graph structure [63]”
From Challenges in AI-driven Biomedical Multimodal Data Fusion and Analysis (2025) · cites this paper as [63]
In the citing paper
“One of the main advantages of using the GNN for processing MTS is its ability to incorporate the information represented by vertices and edges into the learning process.”
From TVGeAN: Tensor Visibility Graph-Enhanced Attention Network for Versatile Multivariant Time Series Learning Tasks (2024) · cites this paper as [26]
2 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Lecture Notes in Computer Science 1205
(1999) · Industrial Robot: An International Journal · doi:10.1108/ir.1999.04926fae.001 · Cited by 38,662
Neural Networks: A Comprehensive Foundation
Haykin, S. (1998) · Cited by 30,070
Wu, Y. (1999) · Technometrics · doi:10.1080/00401706.1999.10485951 · Cited by 27,221 · cited 1×
“Those techniques include random fields [42], Bayesian networks [43], statistical relational learning [44], transductive learning [45], and semisu-pervised approaches for graph processing [46].”
Neural networks and physical systems with emergent collective computational abilities.
Hopfield, J. J. (1982) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.79.8.2554 · Cited by 19,270 · cited 1×
“[47]–[50] and Hopfield neural networks [51].”
Neural networks for pattern recognition
(1994) · Choice Reviews Online · doi:10.5860/choice.31-5500 · Cited by 18,538
The anatomy of a large-scale hypertextual Web search engine
Brin, S., & Page, L. (1998) · Computer Networks and ISDN Systems · doi:10.1016/s0169-7552(98)00110-x · Cited by 16,004 · cited 8×
“More recently, there have been various approaches [17], [18] attempting to preserve the graph structured”
InfluentialParallel Distributed Processing: Explorations in the Microstructure of Cognition: Foundations
Rumelhart, D. E., McClelland, J. L., & AU (1986) · doi:10.7551/mitpress/5236.001.0001 · Cited by 15,686
Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
Lafferty, J., McCallum, A., & Pereira, F. C. N. (2001) · ScholarlyCommons (University of Pennsylvania) · Cited by 12,935
Neural Networks for Pattern Recognition
Bishop, C. M. (1995) · doi:10.1093/oso/9780198538493.001.0001 · Cited by 12,287
Garreau, G. A., & Rudin, W. (1987) · The Statistician · doi:10.2307/2348852 · Cited by 10,622
Research works printed but not linked: 13
- bunke, 2000, graph matching: theoretical foundations, algorithms, and applications, Proc Vis Interface, 82
- fahlman, 1989, Advances in neural information processing systems, 2, 524
- schmitt, 1998, relating chemical structure to activity: an application of the neural folding architecture, Proc Workshop Fuzzy-Neuro Syst /Conf Eng Appl Neural Netw, 170
- k�chler, 1996, Lecture Notes in Computer Science, 1137
- wo?nica, 2006, matching based kernels for labeled graphs, Proc Int Workshop Mining Learn Graphs/ECML/PKDD, 97
- 2005, Proc Int Workshop Sub-Symbol Paradigms Structured Domains
- 2005, Proc Int l Workshop Mining Graphs Trees and Sequences
- 2006, Proc Int l Workshop Mining Graphs Trees and Sequences
- srinivasan, 1994, mutagenesis: ilp experiments in a non-determinate biological domain, Proc 9th Int l Workshop Inductive Logic Programming, 217
- scarselli, 0, The GNN Toolbox
- bua, 2002, Lecture Notes in Computer Science, 2415
- kirsten, 2002, Multirelational distance-based clustering
- 2004, Proc Int Workshop Statist Relat Learn Connect Other Fields
Cited by
Items 801–900 of 1,052
Deep Graph Matching via Blackbox Differentiation of Combinatorial Solvers
Rolínek, M., Swoboda, P., Zietlow, D., Paulus, A., Musil, V., & Martius, G. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58604-1_25 · Cited by 130
Knowledge graph enhanced neural collaborative recommendation
Sang, L., Xu, M. G., Qian, S., & Wu, X. (2020) · Expert Systems with Applications · doi:10.1016/j.eswa.2020.113992 · Cited by 91
Adversarial Active Learning Based Heterogeneous Graph Neural Network for Fake News Detection
Ren, Y., Wang, B., Zhang, J., & Chang, Y. (2020) · Proceedings · doi:10.1109/icdm50108.2020.00054 · Cited by 113
“Graph Neural Networks (GNNs) learn nodes’ new feature vectors through a recursive neighborhood aggregation scheme [12], [33], [46].”
Building Semantic Knowledge Graphs from (Semi-)Structured Data: A Review
Ryen, V., Soylu, A., & Roman, D. (2022) · Future Internet · doi:10.3390/fi14050129 · Cited by 104
“[14], while the graph neural network approaches model a neural network based on the topology of the knowledge graph [15].”
BackgroundSurvey on Machine Learning for Traffic-Driven Service Provisioning in Optical Networks
Panayiotou, T., Michalopoulou, M., & Ellinas, G. (2023) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2023.3247842 · Cited by 94
“They were introduced in [153] and numerous variants, with diverse learning tasks, have been developed since, including feature inference of ver-”
BackgroundCausal-Trivial Attention Graph Neural Network for Fault Diagnosis of Complex Industrial Processes
Wang, H., Liu, R., Ding, S. X., Hu, Q., Li, Z., & Zhou, H. (2023) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2023.3282979 · Cited by 94
“Graph neural networks (GNN) [13] algorithms applied in various fields have shown excellent graph recognition performance and can be used to learn error-oriented graph representations, which is an efficient way to fuse information from multiple components through a message passing mechanism.”
BackgroundArchitectural spatial layout planning using artificial intelligence
Ko, J., Ennemoser, B., Yoo, W.-J., Yan, W., & Clayton, M. J. (2023) · Automation in Construction · doi:10.1016/j.autcon.2023.105019 · Cited by 102
Language-Agnostic Representation Learning of Source Code from Structure and Context
Zügner, D., Kirschstein, T., Catasta, M., Leskovec, J., & Günnemann, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.11318 · Cited by 60
“(Gori et al., 2005; Scarselli et al., 2008) are able to learn vector representations of nodes and graphs in an end-to-end fashion, encoding structural and feature information in the embedding space.”
BackgroundHybrid Graph Neural Networks for Crowd Counting
Luo, A., Yang, F., Li, X. S., Nie, D., Jiao, Z., Zhou, S., & Cheng, H. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i07.6839 · Cited by 87
Multimedia Intelligence: When Multimedia Meets Artificial Intelligence
Zhu, W., Wang, X. Y., & Gao, W. (2020) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2020.2969791 · Cited by 95
“These factors are also taken into account when designing deep neural networks by means of adopting Graph Neural Network [101] or Relation Network [102], [103].”
BackgroundMethodLSTM variants meet graph neural networks for road speed prediction
Lu, Z., Lv, W., Cao, Y., Xie, Z., Peng, H. M., & Du, B. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.03.031 · Cited by 98
ASFGNN: Automated separated-federated graph neural network
Zheng, L., Zhou, J., Chen, C., Wu, B., Wang, L., & Zhang, B. (2021) · Peer-to-Peer Networking and Applications · doi:10.1007/s12083-021-01074-w · Cited by 98
Luo, Y., Lü, C., Zhu, L., & Song, J. (2021) · International Journal of Electrical Power & Energy Systems · doi:10.1016/j.ijepes.2020.106753 · Cited by 87
Recommending metamodel concepts during modeling activities with pre-trained language models
Weyssow, M., Sahraoui, H. A., & Syriani, E. (2022) · Software & Systems Modeling · doi:10.1007/s10270-022-00975-5 · Cited by 105
“Nevertheless, other model architectures, such as graph neural networks based on graph-structured representation of the input, might be a relevant alternative to better learn from the structural aspect of the metamodels [21,36].”
BackgroundText analysis using deep neural networks in digital humanities and information science
Suissa, O., Elmalech, A., & Zhitomirsky‐Geffet, M. (2021) · Journal of the Association for Information Science and Technology · doi:10.1002/asi.24544 · Cited by 109
“It should be noted that there are graph-based DNN models (e.g., Scarselli, Gori, Tsoi, Hagenbuchner, & Monfardini, 2008), but they are not suitable for the NER task.”
Xing, L., & Perelman, L. (2022) · Journal of Water Resources Planning and Management · doi:10.1061/(asce)wr.1943-5452.0001550 · Cited by 91
“Theconcept of GNNs was first developed to extend existing NNs for processing the data represented in graph domains (Scarselli et al. 2009).”
Spatio-temporal Convolutional Network Based Power Forecasting of Multiple Wind Farms
Dong, X., Sun, Y., Li, Y., Wang, X., & Pu, T. (2022) · Journal of Modern Power Systems and Clean Energy · doi:10.35833/mpce.2020.000849 · Cited by 94
Sun, L., Rao, Y., Zhang, X., Lan, Y., & Yu, S. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i4.20334 · Cited by 87
“Compared with the traditional GNNs, our proposed sequential HGATs can depict the dynamic interactions be-The tween users and cascades through hyperedges and attention mechanism.”
BackgroundBereczki, Z., Benczik, B., Balogh, O. M., Marton, S., Puhl, E., Pétervári, M., Váczy‐Földi, M., Papp, Z. T., Makkos, A., Glass, K., et al. (2024) · British Journal of Pharmacology · doi:10.1111/bph.17302 · Cited by 94
“The subclass of deep neural network models that are designed to handle graph data is called graph neural networks (GNNs) (Gori et al., 2005; Scarselli et al., 2009).”
Hypergraph-Induced Convolutional Networks for Visual Classification
Shi, H., Zhang, Y., Zhang, Z., Ma, N., Zhao, X., Gao, Y., & Sun, J. (2018) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2869747 · Cited by 92
“The graph-structured neural network with the form of recurrent neural networks (RNNs) was proposed in [30] and [31].”
Amortized Causal Discovery: Learning to Infer Causal Graphs from Time-Series Data
Löwe, S., Madras, D., Zemel, R. S., & Welling, M. (2020) · Data Archiving and Networked Services (DANS) · doi:10.48550/arxiv.2006.10833 · Cited by 64
“As a result, our loss function L is a variational lower bound: Encoder The encoder q φ ( z | x ) applies a graph neural network f enc ,φ (Scarselli et al., 2008; Li et al., 2016; Gilmer et al., 2017; Kipf and Welling, 2017) to the input, which propagates information across a fully connected graph G…”
BackgroundMethodVision-Based Freezing of Gait Detection With Anatomic Directed Graph Representation
Hu, K., Wang, Z., Mei, S., Martens, K. A. E., Yao, T., Lewis, S. J. G., & Feng, D. D. (2019) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2019.2923209 · Cited by 91
“The recently proposed GCNN is appropriate to take these FoG related knowledge into consideration and to de-sign an effective fine-grained human action recognition method for FoG detection [20]–[23], [36].”
MethodSTEP: A Spatio-Temporal Fine-Granular User Traffic Prediction System for Cellular Networks
Yu, L., Li, M., Jin, W., Guo, Y., Wang, Q., Yan, F., & Li, P. (2020) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2020.3001225 · Cited by 92
“Feasible approaches include GCN [48], [50], [51], [52], GNN [53], [54], [55] and spectral networks [56].”
MethodParameter-Efficient Person Re-Identification in the 3D Space
Zheng, Z., Wang, X., Zheng, N., & Yang, Y. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3214834 · Cited by 107
“To address the limitation, we follow the idea of GNNs [80] to build the graph G based on the distance between points.”
BackgroundMethodResGNet-C: A graph convolutional neural network for detection of COVID-19
Yu, X., Lu, S., Guo, L., Wang, S., & Zhang, Y. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.07.144 · Cited by 98
“where m is the dimension of Euclidean space [35].”
Ouyang, S., & Li, Y. (2020) · Remote Sensing · doi:10.3390/rs13010119 · Cited by 91
“Due to the irregular distribution of ground objects and connections between objects only in mutual relations, the objects and their relationships form a graph, where graph nodes represent objects and the connecting edges between the graph nodes denote the spatial relationship between the objects [45], such as neighbor…”
BackgroundHuang, J., Gao, L., Li, X., & Zhang, C. (2023) · Journal of Manufacturing Systems · doi:10.1016/j.jmsy.2023.06.007 · Cited by 91
Zhang, J., Xiao, F., Li, A., Ma, T., Xu, K., Zhang, H., Yan, R., Fang, X., Li, Y., & Wang, D. (2023) · Building and Environment · doi:10.1016/j.buildenv.2023.110600 · Cited by 90
Query performance prediction for concurrent queries using graph embedding
Zhou, X., Sun, J., Li, G., & Feng, J. (2020) · Proceedings of the VLDB Endowment · doi:10.14778/3397230.3397238 · Cited by 86
“Primitive GNNs [22] only support static graphs with fixed vertices and have to re-train when graph changes.”
BackgroundPermutation Invariant Graph Generation via Score-Based Generative ModelingPreprint
Niu, C., Song, Y. S., Song, J., Zhao, S., Grover, A., & Ermon, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.00638 · Cited by 304
“These models typically use a graph neural network (GNN) (Gori et al., 2005; Scarselli et al., 2008) to encode graph data to a latent space, and generate samples by decoding latent variables sampled from a prior distribution.”
BackgroundMethodLearning Edge Representations via Low-Rank Asymmetric Projections
Abu-El-Haija, S., Perozzi, B., & Al-Rfou, R. (2017) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3132847.3132959 · Cited by 95
Tang, J., Han, X., Pan, J., Jia, K., & Tong, X. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00467 · Cited by 101
“A GCNN is finally trained to learn the mesh deformation for producing geometric details.”
MethodInfluentialProbing stop pair production at the LHC with graph neural networks
Abdughani, M., Ren, J., Wu, L., & Yang, J. M. (2019) · Journal of High Energy Physics · doi:10.1007/jhep08(2019)055 · Cited by 85
Graph Sequence Recurrent Neural Network for Vision-Based Freezing of Gait Detection
Hu, K., Wang, Z., Wang, W., Martens, K. A. E., Wang, L., Tan, T. N., Lewis, S. J. G., & Feng, D. D. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2946469 · Cited by 91
“Graph neural network (GNN) [32] was proposed as one of the first graph-based neural networks to process the data containing graph structures.”
Syntax-Aware Graph Attention Network for Aspect-Level Sentiment Classification
Huang, L., Sun, X., Li, S., Zhang, L., & Wang, H. (2020) · International Conference on Computational Linguistics · doi:10.18653/v1/2020.coling-main.69 · Cited by 104
“GNN was first proposed in (Scarselli et al., 2009) and has been used in many tasks in NLP including text classification (Defferrard et al., 2016), sequence labeling (Zhang et al., 2018), neural machine translation (Bastings et al., 2017), and relational reasoning (Battaglia et al., 2016).”
MethodZhang, Y., Xia, J., & Jiang, B. (2022) · The Journal of Chemical Physics · doi:10.1063/5.0080766 · Cited by 88
Short-Term Wind Power Prediction via Spatial Temporal Analysis and Deep Residual Networks
Li, H. (2022) · Frontiers in Energy Research · doi:10.3389/fenrg.2022.920407 · Cited by 85
“As we all know, a graph is a kind of structured data, which comprises a series of objects (nodes) and relationship types (edges) (Scarselli et al., 2008).”
BackgroundGouk, D., Kwon, M., Bae, H., Lee, S., & Jung, M. (2023) · IEEE Micro · doi:10.1109/mm.2023.3237491 · Cited by 85
“Graph neural networks (GNNs) 9 can also take the bene fi ts from a large storage capacity that memory pooling offers.”
BackgroundGraph Neural Networks (GNNs) based accelerated numerical simulation
Jiang, C., & Chen, N.-Z. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.106370 · Cited by 95
Vatter, J., Mayer, R., & Jacobsen, H. (2023) · ACM Computing Surveys · doi:10.1145/3597428 · Cited by 113
“Therefore, a new type of Deep Neural Network (DNN) architecture was developed in the late 2000s, the so-called Graph Neural Network (GNN) [60, 155].”
MethodAir traffic density prediction using Bayesian ensemble graph attention network (BEGAN)
Xu, Q., Pang, Y., & Liu, Y. (2023) · Transportation Research Part C Emerging Technologies · doi:10.1016/j.trc.2023.104225 · Cited by 92
HDFormer: High-order Directed Transformer for 3D Human Pose Estimation
Chen, H., He, J.-Y., Xiang, W., Cheng, Z.-Q., Liu, W., Liu, H., Luo, B., Geng, Y., & Xie, X. (2023) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2023/65 · Cited by 85
Wan, X., Li, Z., Yu, W., Wang, A., Ke, X., Guo, H., Su, J., Li, L., Gui, Q., Zhao, S., et al. (2023) · Advanced Materials · doi:10.1002/adma.202305192 · Cited by 92
Zhao, Y., Luo, X., & Zhang, Y. (2023) · Computers & Industrial Engineering · doi:10.1016/j.cie.2023.109802 · Cited by 88
Constrained multi-objective optimization problems: Methodologies, algorithms and applications
Hao, Y., Zhao, C., Zhang, Y., Cao, Y., & Li, Z. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.111998 · Cited by 96
Zhu, S., Sarıtürk, ., & Arróyave, R. (2025) · Acta Materialia · doi:10.1016/j.actamat.2025.120747 · Cited by 97
SceneGraphNet: Neural Message Passing for 3D Indoor Scene Augmentation
Zhou, Y., While, Z., & Kalogerakis, E. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00748 · Cited by 79
Exact and Inexact Graph Matching: Methodology and Applications
Riesen, K., Jiang, X., & Horst Bunke, A. (2010) · The Kluwer international series on advances in database systems · doi:10.1007/978-1-4419-6045-0_7 · Cited by 100
“In [52, 72] the graph neural network is crucially extended such that also undirected and acyclic graphs can be processed.”
BackgroundEcho state graph neural networks with analogue random resistive memory arrays
Wang, S., Li, Y., Wang, D., Zhang, W., Chen, X., Dong, D., Wang, S., Zhang, X., Lin, P., Gallicchio, C., et al. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00609-5 · Cited by 84
“s() j = as (t) j + (1 − a)σ [uj + ∑ k∈N(j) WRs k ] G (1)”
MethodWang, X., Fu, L., Cheng, N., Sun, R., Luan, T. H., Quan, W., & Aldubaikhy, K. (2022) · Remote Sensing · doi:10.3390/rs14174377 · Cited by 99
Khadivi, M., Charter, T., Yaghoubi, M., Jalayer, M., Ahang, M., Shojaeinasab, A., & Najjaran, H. (2025) · Computers & Industrial Engineering · doi:10.1016/j.cie.2025.110856 · Cited by 105
Perceiver CPI: a nested cross-attention network for compound–protein interaction prediction
Nguyen, N.-Q., Jang, G., Kim, H., & Kang, J. (2023) · Bioinformatics · doi:10.1093/bioinformatics/btac731 · Cited by 83
Bridging the Structural Gap Between Encoding and Decoding for Data-To-Text Generation
Zhao, C., Walker, M., & Chaturvedi, S. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.224 · Cited by 83
“Graph neural networks (GNN) (Scarselli et al., 2009) aim to learn a latent state representation for each node in a graph by aggregating local information from its neighbors and the connected edges.”
BackgroundHierarchical Human Semantic Parsing with Comprehensive Part-Relation Modeling
Wang, W., Zhou, T., Qi, S., Shen, J., & Zhu, S. C. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2021.3055780 · Cited by 91
“For such problems, approximate iterative inference algorithms are often adopted [40], [41].”
BackgroundMethodInfluentialLime: Low-Cost and Incremental Learning for Dynamic Heterogeneous Information Networks
Peng, H. M., Yang, R., Wang, Z., Li, J., He, L., Yu, P. S., Zomaya, A. Y., & Ranjan, R. K. (2021) · IEEE Transactions on Computers · doi:10.1109/tc.2021.3057082 · Cited by 92
“ods [15], [16], are primarily concerned about static networks,”
BackgroundThe Logic of Graph Neural Networks
Grohe, M. (2021) · ACM/IEEE Symposium on Logic in Computer Science (LICS) · doi:10.1109/lics52264.2021.9470677 · Cited by 167
“Many different versions of the GNNs have been considered in the literature (see, for example, [18, 19, 23, 32, 50]).”
BackgroundLorentzian Graph Convolutional Networks
Zhang, Y., Wang, X., Shi, C., Liu, N., & Song, G. (2021) · International World Wide Web Conference (WWW) · doi:10.1145/3442381.3449872 · Cited by 128
“Graph neural networks [15, 45], which extend the deep neural network to deal with graph data, have achieved great success in solving machine learning problems.”
BackgroundAttentive sequential model based on graph neural network for next poi recommendation
Wang, D., Wang, X., Xiang, Z., Yu, D., Deng, S., & Xu, G. (2021) · World Wide Web · doi:10.1007/s11280-021-00961-9 · Cited by 85
“For example, Scarselli et al. [32] propose a GNN based model with information propagation mechanism.”
BackgroundApplication of message passing neural networks for molecular property prediction
Tang, M., Li, B., & Chen, H. (2023) · Current Opinion in Structural Biology · doi:10.1016/j.sbi.2023.102616 · Cited by 102
Next POI Recommendation with Dynamic Graph and Explicit Dependency
Yin, F., Liu, Y., Shen, Z., Chen, L., Shang, S., & Han, P. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i4.25608 · Cited by 113
“As the check-in data for next POI recommendation can be simply described by sparse graphs, this property motivates the application of GNN for POI recommendation.”
MethodInfluentialSQL-to-Text Generation with Graph-to-Sequence Model
Xu, K., Wu, L., Wang, Z., Feng, Y., & Sheinin, V. (2018) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d18-1112 · Cited by 101
“Following (Scarselli et al., 2009), this method adds a super node vs that is connected to all other nodes by a special type of edge.”
MethodLemos, H., Prates, M., Avelar, P., & Lamb, L. C. (2019) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai.2019.00125 · Cited by 115
“We also formalise our solution using a Graph Neural Network (GNN) framework [14], [16], [18].”
MethodThe More You Know: Using Knowledge Graphs for Image ClassificationPreprint
Marino, K., Salakhutdinov, R., & Gupta, A. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1612.04844 · Cited by 99
“This paper (an extension of Graph Neural Networks [31]) serves as the foundation for our Graph Search Neural Network (GSNN).”
BackgroundMethodInfluentialInductive Relation Prediction by Subgraph ReasoningPreprint
Teru, K. K., Denis, E., & Hamilton, W. L. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.06962 · Cited by 111
“We present a Graph Neural Network (GNN) (Scarselli et al., 2008; Bronstein et al., 2017) framework (GraIL: Graph Inductive Learning) that has a strong inductive bias to learn entity-independent relational semantics.”
MethodGallicchio, C., & Micheli, A. (2012) · Neurocomputing · doi:10.1016/j.neucom.2012.08.017 · Cited by 99
AGCN: Attention-based graph convolutional networks for drug-drug interaction extraction
Park, C., Park, J., & Park, S. (2020) · Expert Systems with Applications · doi:10.1016/j.eswa.2020.113538 · Cited by 89
“In early works (Gori, Monfardini, & Scarselli, 2005; Scarselli, Gori, Tsoi, Hagenbuchner, & Monfardini, 2009), researchers proposed a type of graph neural networks (GNN) to address more general graphs, such as cyclic, undirected, and directed graphs.”
BackgroundFKAConv: Feature-Kernel Alignment for Point Cloud Convolution
Boulch, A., Puy, G., & Marlet, R. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-69525-5_23 · Cited by 85
Deep learning in the COVID-19 epidemic: A deep model for urban traffic revitalization index
Lv, Z., Li, J., Dong, C., Li, H., & Xu, Z. (2021) · Data & Knowledge Engineering · doi:10.1016/j.datak.2021.101912 · Cited by 84
“Compared with he most basic structure (Multi-Layer Perception, MLP) of neural network, graph neural network [14] adds an adjacency matrix to articipate in the process of matrix calculation, as shown in formula (1).”
Capturing the objects of vision with neural networks
Peters, B., & Kriegeskorte, N. (2021) · Nature Human Behaviour · doi:10.1038/s41562-021-01194-6 · Cited by 75
On the Reproducibility and Replicability of Deep Learning in Software Engineering
Liu, C., Gao, C., Xia, X., Lo, D., Grundy, J., & Yang, X. (2022) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3477535 · Cited by 61
Learning Graph Representations with Embedding PropagationPreprint
García-Durán, A., & Niepert, M. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1710.03059 · Cited by 88
“Graph neural networks (GNN) [39] is a general class of recursive neural networks for graphs where each node is associated with one label.”
BackgroundEquivariant Subgraph Aggregation NetworksPreprint
Bevilacqua, B., Frasca, F., Lim, D., Srinivasan, B., Cai, C., Balamurugan, G., Bronstein, M. M., & Maron, H. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.02910 · Cited by 195
“Their roots can be traced back to the field of computational chemistry Kireev (1995); Baskin et al. (1997) and the pioneering works of Sperduti (1994); Goller & Kuchler (1996); Sperduti & Starita (1997); Gori et al. (2005); Scarselli et al. (2008).”
BackgroundLearning to Represent the Evolution of Dynamic Graphs with Recurrent Models
Taheri, A., Gimpel, K., & Berger-Wolf, T. Y. (2019) · International World Wide Web Conference (WWW) · doi:10.1145/3308560.3316581 · Cited by 106
“Graph Neural Networks (GNNs) [14, 33] are a well-known neural network architecture for processing graph structures, and the main idea is to learn the topology of a graph via an iterative propagation procedure.”
BackgroundClassification of Multiple Sclerosis Clinical Profiles via Graph Convolutional Neural Networks
Marzullo, A., Kocevar, G., Stamile, C., Durand‐Dubief, F., Terracina, G., Calimeri, F., & Sappey‐Marinier, D. (2019) · Frontiers in Neuroscience · doi:10.3389/fnins.2019.00594 · Cited by 82
“Since the first definition of Graph Neural Networks (GNN) (Scarselli et al., 2009), a huge effort was made to extend neural networks with the purpose of processing graph structures data.”
BackgroundReducing Communication in Graph Neural Network Training
Tripathy, A., Yelick, K., & Buluç, A. (2020) · International Conference for High Performance Computing, Networking, Storage and Analysis (SC) · doi:10.1109/sc41405.2020.00074 · Cited by 134
“Graph Neural Networks (GNNs) [25] are types of neural networks that use the connectivity information that is natural in datasets that can be represented as graphs, such as molecules, transportation and social networks, the power grid, and proteins.”
BackgroundData Analytics on Graphs Part III: Machine Learning on Graphs, from Graph Topology to Applications
Stanković, L., Mandic, D. P., Daković, M., Brajović, M., Dees, B. S., Li, S., & Constantinides, A. G. (2020) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000078-3 · Cited by 99
Collective variable-based enhanced sampling and machine learning
Chen, M. (2021) · The European Physical Journal B · doi:10.1140/epjb/s10051-021-00220-w · Cited by 84
Multitask Hypergraph Convolutional Networks: A Heterogeneous Traffic Prediction Framework
Wang, J., Zhang, Y., Wang, L., Hu, Y., Piao, X., & Yin, B. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3168879 · Cited by 89
“Due to the natural graph topology of traffic data, graph-based neural networks [5] such as Spatio-temporal Graph Convolutional Networks (STGCN) [6] and Diffusion Convolutional Recurrent Neural Network (DCRNN) [7] demonstrate the remarkable role of spatio-temporal feature extraction.”
BackgroundShort-term residential load forecasting using Graph Convolutional Recurrent Neural Networks
Arastehfar, S., Matinkia, M., & Jabbarpour, M. R. (2022) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2022.105358 · Cited by 84
Machine learning prediction of structural dynamic responses using graph neural networks
Li, Q., Wang, Z., Li, L., Hao, H., Chen, W., & Shao, Y. (2023) · Computers & Structures · doi:10.1016/j.compstruc.2023.107188 · Cited by 97
GDDN: Graph Domain Disentanglement Network for Generalizable EEG Emotion Recognition
Chen, B., Chen, C. L. P., & Zhang, T. (2024) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2024.3371540 · Cited by 92
“The early literature [39] on GNNs learned a converged static state representation of a target node with a transition function adopted to its neighbor.”
BackgroundAI meets physics: a comprehensive survey
Jiao, L., Song, X., You, C., Liu, X., Li, L., Chen, P., Tang, X., Feng, Z., Liu, F., Guo, Y., et al. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10874-4 · Cited by 99
AI Methods for Antimicrobial Peptides: Progress and Challenges
Brizuela, C. A., Liu, G., Stokes, J. M., & de la Fuente‐Núñez, C. (2025) · Microbial Biotechnology · doi:10.1111/1751-7915.70072 · Cited by 87
“Graph neural networks (Scarselli et al. 2009) have emerged as powerful learning tools to capture complex interrelations among components in various subjects of study.”
A Unified Lottery Ticket Hypothesis for Graph Neural NetworksPreprint
Chen, T., Sui, Y., Chen, X., Zhang, A., & Wang, Z. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.06790 · Cited by 173
“There are mainly three categories of GNNs (Dwivedi et al., 2020): i) extending original convolutional neural networks to the graph regime (Scarselli et al., 2008; Bruna et al., 2013; Kipf & Welling, 2016; Hamilton et al., 2017); ii) introducing anisotropic operations on graphs such as gating and…”
BackgroundTransferability of Spectral Graph Convolutional Neural NetworksPreprint
Levie, R., Huang, W., Bucci, L., Bronstein, M. M., & Kutyniok, G. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.12972 · Cited by 179
“Some popular examples of spatial methods are [7, 12, 10].”
BackgroundTime-Series Graph Network for Sea Surface Temperature Prediction
Sun, Y., Yao, X., Bi, X., Huang, X., Zhao, X., & Qiao, B. (2021) · Big Data Research · doi:10.1016/j.bdr.2021.100237 · Cited by 83
Adaptive spatial-temporal graph attention networks for traffic flow forecasting
Kong, X., Zhang, J., Wei, X., Xing, W., & Lu, W. B. (2021) · Applied Intelligence · doi:10.1007/s10489-021-02648-0 · Cited by 93
“Graph Neural Networks (GNNs) was first mentioned in [46], and [47] can handle more general graphs directly, including cyclic, directed, and undirected graphs [48].”
BackgroundZuo, Z., Wang, P., Chen, X., Li, T., Ge, H., & Qian, D. (2021) · BMC Bioinformatics · doi:10.1186/s12859-021-04352-9 · Cited by 88
Zola, F., Segurola-Gil, L., Bruse, J. L., Galar, M., & Orduna-Urrutia, R. (2022) · Computers & Security · doi:10.1016/j.cose.2022.102632 · Cited by 84
RouteNet-Erlang: A Graph Neural Network for Network Performance Evaluation
Ferriol-Galmés, M., Rusek, K., Suárez‐Varela, J., Xiao, S., Shi, X., Cheng, X., Wu, B., Barlet‐Ros, P., & Cabellos‐Aparicio, A. (2022) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom48880.2022.9796944 · Cited by 115
“In order to construct positive interarrival times, z is mapped by a nonlinear transformation: δt = F −1 E ( λ, FN ( 0, s(2), zt )) , where FN (0, s(2), ·) and FE(λ, ·) are respectively a CDF of the normal distribution with μ = 0 and variance s(2)=[3, 12], and an exponential distribution with parameter λ=[40, 2000].”
BackgroundMethodInfluentialCell graph neural networks enable the precise prediction of patient survival in gastric cancer
Wang, Y., Wang, Y. G., Hu, C., Li, M., Fan, Y., Otter, N., Sam, I., Gou, H., Hu, Y., Kwok, T., et al. (2022) · npj Precision Oncology · doi:10.1038/s41698-022-00285-5 · Cited by 71
AMGNET: multi-scale graph neural networks for flow field prediction
Yang, Z., Dong, Y., Deng, X., & Zhang, L. (2022) · Connection Science · doi:10.1080/09540091.2022.2131737 · Cited by 88
“Graph neural networks were first proposed by Gori et al. (2005) and Scarselli et al. (2009).”
MethodSMG: A Micro-gesture Dataset Towards Spontaneous Body Gestures for Emotional Stress State Analysis
Chen, H., Shi, H., Liu, X., Li, X., & Zhao, G. (2023) · International Journal of Computer Vision · doi:10.1007/s11263-023-01761-6 · Cited by 128
“Limited sample amount easily leads to over-fitting on these models (Scarselli et al. 2008) (e.g., in our cases, a graph with ) are suitable for insuffi-cient samples to get node “gradients” without the need for high-dimensional weights.”
BackgroundHusformer : A Multimodal Transformer for Multimodal Human State Recognition
Wang, R., Jo, W., Zhao, D., Wang, W., Gupte, A., Yang, B., Chen, G., & Min, B. (2024) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2024.3357618 · Cited by 87
“Moreover, to comprehensively evaluate if our Husformer could improve the performance compared with using single modality, we also implemented Long Short-Term Memory Nerual Network (LSTM) [52], Graph Neural Network (GNN) [53] and Transformer [41], which were broadly utilized for single-modal-based human state recogniti…”
MethodE-GRACL: an IoT intrusion detection system based on graph neural networks
Lin, L., Zhong, Q., Qiu, J., & Liang, Z. (2024) · The Journal of Supercomputing · doi:10.1007/s11227-024-06471-5 · Cited by 97
Machine learning-based mechanical performance prediction and design of lattice structures
Liu, Y., Huang, W., Wang, Z., Zhang, J., & Liu, J. (2025) · International Journal of Mechanical Sciences · doi:10.1016/j.ijmecsci.2025.110230 · Cited by 80
Towards the unification of structural and statistical pattern recognition
Horst Bunke, A., & Riesen, K. (2011) · Pattern Recognition Letters · doi:10.1016/j.patrec.2011.04.017 · Cited by 91
Modeling Semantics with Gated Graph Neural Networks for Knowledge Base Question Answering
Sorokin, D., & Gurevych, I. (2018) · TUbilio (Technical University of Darmstadt) · doi:10.48550/arxiv.1808.04126 · Cited by 116
“Li et al. (2016) have developed the gated architecture based on the graph neural network formulation of Scarselli et al. (2009).”
BackgroundGraph Neural Networks for Particle Tracking and Reconstruction
Duarte, J. M., & Vlimant, J.-R. (2022) · WORLD SCIENTIFIC eBooks · doi:10.1142/9789811234026_0012 · Cited by 95
On the Equivalence between Positional Node Embeddings and Structural Graph RepresentationsPreprint
Srinivasan, B., & Ribeiro, B. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.00452 · Cited by 106
“Graph neural networks (GNNs) (e.g., (Duvenaud et al., 2015; Gilmer et al., 2017; Kipf & Welling, 2016a; Hamilton et al., 2017a; Xu et al., 2018; Scarselli et al., 2008) among others) exploit this approach in tasks such as node and whole-graph classification.”
Metadata
- Authors
- Scarselli, F., Gori, M., Tsoi, A. C., Hagenbuchner, M., & Monfardini, G.
- Affiliations
- University of Siena (IT)
- Hong Kong Baptist University (HK)
- University of Wollongong (AU)
- Venue
- IEEE Transactions on Neural Networks
- Published
- 2009-01
- Volume
- 20
- Issue
- 1
- Pages
- 61-80
- ISSN
- 1045-9227 (ISSN-L)
- 1941-0093
- OpenAlex
- W2116341502
- Cited by
- 9,582
- References
- 111
- Citations by year
- 2026: 1235 · 2025: 1651 · 2024: 1503 · 2023: 1398 · 2022: 1177 · 2021: 1150 · 2020: 822 · 2019: 375 · 2018: 149 · 2017: 44 · 2016: 15 · 2015: 10 · 2014: 10 · 2013: 12 · 2012: 6
- FWCI
- 13.19
- Type
- journal article
- Language
- en
- Open access
- green
- figshare.com
- Locations
- IEEE Transactions on Neural Networks (publishedVersion)
- PubMed (publishedVersion)
- HKBU Institutional Repository (Hong Kong Baptist University) (submittedVersion)
- Research Online (University of Wollongong) (submittedVersion)
- Figshare (submittedVersion, cc-by)
- Use Siena air (University of Siena) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Advanced Graph Neural Networks
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Graph Theory and Algorithms
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Keywords
- graph neural networks, graph-structured data, node embeddings, supervised learning algorithms, parameter estimation, computational cost, generalization, directed and undirected graphs
- MeSH terms
- Algorithms, Artificial Intelligence, Pattern Recognition, Automated, Regression Analysis, Reproducibility of Results, Linear Models, Databases, Factual, Neural Networks, Computer, Nonlinear Dynamics, Internet
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.