GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing
In several areas of science and engineering, data can be naturally represented in graph form, where nodes denote entities and edges stand for relationships between them. Graph Neural Networks (GNNs) are a well-known class of machine learning models for graph processing. In this paper, we present GNN...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-06-01
|
| Series: | SoftwareX |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352711022000486 |
| _version_ | 1828318354596691968 |
|---|---|
| author | Niccolò Pancino Pietro Bongini Franco Scarselli Monica Bianchini |
| author_facet | Niccolò Pancino Pietro Bongini Franco Scarselli Monica Bianchini |
| author_sort | Niccolò Pancino |
| collection | DOAJ |
| description | In several areas of science and engineering, data can be naturally represented in graph form, where nodes denote entities and edges stand for relationships between them. Graph Neural Networks (GNNs) are a well-known class of machine learning models for graph processing. In this paper, we present GNNkeras, a library, based on Keras, which allows the implementation of a large subclass of GNNs. GNNkeras is a flexible tool: the implemented models can be used to classify/cluster nodes, edges, or whole graphs. Moreover, GNNkeras can be applied to both homogeneous and heterogeneous graphs, exploiting both inductive and mixed inductive–transductive learning, and can implement a layered version of GNNs, namely the LGNN model. |
| first_indexed | 2024-04-13T17:40:46Z |
| format | Article |
| id | doaj.art-a9768e0d7591480ea3bb24666167b2c1 |
| institution | Directory Open Access Journal |
| issn | 2352-7110 |
| language | English |
| last_indexed | 2024-04-13T17:40:46Z |
| publishDate | 2022-06-01 |
| publisher | Elsevier |
| record_format | Article |
| series | SoftwareX |
| spelling | doaj.art-a9768e0d7591480ea3bb24666167b2c12022-12-22T02:37:12ZengElsevierSoftwareX2352-71102022-06-0118101061GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processingNiccolò Pancino0Pietro Bongini1Franco Scarselli2Monica Bianchini3University of Siena, Department of Information Engineering and Mathematics, Via Roma 56, 53100, Siena (SI), Italy; University of Florence, Department of Information Engineering, Via S. Marta 3, 50139, Florence (FI), Italy; Corresponding author at: University of Siena, Department of Information Engineering and Mathematics, Via Roma 56, 53100, Siena (SI), Italy.University of Siena, Department of Information Engineering and Mathematics, Via Roma 56, 53100, Siena (SI), Italy; University of Florence, Department of Information Engineering, Via S. Marta 3, 50139, Florence (FI), ItalyUniversity of Siena, Department of Information Engineering and Mathematics, Via Roma 56, 53100, Siena (SI), ItalyUniversity of Siena, Department of Information Engineering and Mathematics, Via Roma 56, 53100, Siena (SI), ItalyIn several areas of science and engineering, data can be naturally represented in graph form, where nodes denote entities and edges stand for relationships between them. Graph Neural Networks (GNNs) are a well-known class of machine learning models for graph processing. In this paper, we present GNNkeras, a library, based on Keras, which allows the implementation of a large subclass of GNNs. GNNkeras is a flexible tool: the implemented models can be used to classify/cluster nodes, edges, or whole graphs. Moreover, GNNkeras can be applied to both homogeneous and heterogeneous graphs, exploiting both inductive and mixed inductive–transductive learning, and can implement a layered version of GNNs, namely the LGNN model.http://www.sciencedirect.com/science/article/pii/S2352711022000486GraphsMachine LearningTensorFlowKerasGraph Neural Networks |
| spellingShingle | Niccolò Pancino Pietro Bongini Franco Scarselli Monica Bianchini GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing SoftwareX Graphs Machine Learning TensorFlow Keras Graph Neural Networks |
| title | GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing |
| title_full | GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing |
| title_fullStr | GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing |
| title_full_unstemmed | GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing |
| title_short | GNNkeras: A Keras-based library for Graph Neural Networks and homogeneous and heterogeneous graph processing |
| title_sort | gnnkeras a keras based library for graph neural networks and homogeneous and heterogeneous graph processing |
| topic | Graphs Machine Learning TensorFlow Keras Graph Neural Networks |
| url | http://www.sciencedirect.com/science/article/pii/S2352711022000486 |
| work_keys_str_mv | AT niccolopancino gnnkerasakerasbasedlibraryforgraphneuralnetworksandhomogeneousandheterogeneousgraphprocessing AT pietrobongini gnnkerasakerasbasedlibraryforgraphneuralnetworksandhomogeneousandheterogeneousgraphprocessing AT francoscarselli gnnkerasakerasbasedlibraryforgraphneuralnetworksandhomogeneousandheterogeneousgraphprocessing AT monicabianchini gnnkerasakerasbasedlibraryforgraphneuralnetworksandhomogeneousandheterogeneousgraphprocessing |