Deep graph similarity learning: a survey

Ma, Guixiang; Ahmed, Nesreen K.; Willke, Theodore L.; Yu, Philip S.

doi:10.1007/s10618-020-00733-5

Cited by 60 publications

(30 citation statements)

References 97 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, our work relates to recent work on learning distance functions over pairs of graphs (Ma et al, 2019). Recent work (Li et al, 2019), in a model they call Graph Matching Network, has proposed using a cross-graph node-to-node attention approach for solving the problem, and compared it to an approach without cross-graph attention, that is closely related to our models for the Discrimination task.…”

Section: Graph Neural Networkmentioning

confidence: 96%

SpatialSim: Recognizing Spatial Configurations of Objects With Graph Neural Networks

Teodorescu¹,

Hofmann

Oudeyer³

2022

Front. Artif. Intell.

View full text Add to dashboard Cite

An embodied, autonomous agent able to set its own goals has to possess geometrical reasoning abilities for judging whether its goals have been achieved, namely it should be able to identify and discriminate classes of configurations of objects, irrespective of its point of view on the scene. However, this problem has received little attention so far in the deep learning literature. In this paper we make two key contributions. First, we propose SpatialSim (Spatial Similarity), a novel geometrical reasoning diagnostic dataset, and argue that progress on this benchmark would allow for diagnosing more principled approaches to this problem. This benchmark is composed of two tasks: “Identification” and “Discrimination,” each one instantiated in increasing levels of difficulty. Secondly, we validate that relational inductive biases—exhibited by fully-connected message-passing Graph Neural Networks (MPGNNs)—are instrumental to solve those tasks, and show their advantages over less relational baselines such as Deep Sets and unstructured models such as Multi-Layer Perceptrons. We additionally showcase the failure of high-capacity CNNs on the hard Discrimination task. Finally, we highlight the current limits of GNNs in both tasks.

show abstract

Section: Graph Neural Networkmentioning

confidence: 96%

SpatialSim: Recognizing Spatial Configurations of Objects With Graph Neural Networks

Teodorescu¹,

Hofmann

Oudeyer³

2022

Front. Artif. Intell.

View full text Add to dashboard Cite

show abstract

“…Deep graph kernel models have recently emerged, replacing manually designed features with ones learned automatically from data via deep neural networks. According to a recent GSL survey [ 15 ], DDGK [ 19 ] is the only contribution capable of managing heterogeneous and attributed graphs, supporting cross-graph interactions, with already tested applications on chemoinformatics and bioinformatics. Our work still belongs to this group, but focuses on event embeddings.…”

Section: Related Workmentioning

confidence: 99%

“…To solve this task, graph similarity metrics based on predefined distance notions and feature engineering are very costly to compute in practice and hardly generalizable [ 14 ]. In the last few years, researchers have therefore formulated graph similarity estimation as a learning problem [ 15 ]. Deep graph similarity learning (GSL) uses deep strategies to automatically learn a metric for measuring the similarity scores between graph object pairs.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Event Graph Representation and Similarity Learning on Biomedical Literature

Frisoni

Moro

Carlassare

et al. 2021

Sensors

View full text Add to dashboard Cite

The automatic extraction of biomedical events from the scientific literature has drawn keen interest in the last several years, recognizing complex and semantically rich graphical interactions otherwise buried in texts. However, very few works revolve around learning embeddings or similarity metrics for event graphs. This gap leaves biological relations unlinked and prevents the application of machine learning techniques to promote discoveries. Taking advantage of recent deep graph kernel solutions and pre-trained language models, we propose Deep Divergence Event Graph Kernels (DDEGK), an unsupervised inductive method to map events into low-dimensional vectors, preserving their structural and semantic similarities. Unlike most other systems, DDEGK operates at a graph level and does not require task-specific labels, feature engineering, or known correspondences between nodes. To this end, our solution compares events against a small set of anchor ones, trains cross-graph attention networks for drawing pairwise alignments (bolstering interpretability), and employs transformer-based models to encode continuous attributes. Extensive experiments have been done on nine biomedical datasets. We show that our learned event representations can be effectively employed in tasks such as graph classification, clustering, and visualization, also facilitating downstream semantic textual similarity. Empirical results demonstrate that DDEGK significantly outperforms other state-of-the-art methods.

show abstract

“…Related works. So far, several surveys have reviewed deep graph-related approaches such as those mainly focusing on graph representation learning methods [6], [72]- [78], graph attention models [79], knowledge graph research [80], [81], attack and defense techniques on graph data [82], and graph matching approaches [83], [84]. Although most of these surveys have made a passing reference to some modern graph generators, this field requires individual attention due to its value and growing popularity.…”

Section: Introductionmentioning

confidence: 99%

Deep Graph Generators: A Survey

et al. 2021

View full text Add to dashboard Cite

Deep generative models have achieved great success in areas such as image, speech, and natural language processing in the past few years. Thanks to the advances in graph-based deep learning, and in particular graph representation learning, deep graph generation methods have recently emerged with new applications ranging from discovering novel molecular structures to modeling social networks. This paper conducts a comprehensive survey on deep learning-based graph generation approaches and classifies them into five broad categories, namely, autoregressive, autoencoder-based, reinforcement learning-based, adversarial, and flow-based graph generators, providing the readers a detailed description of the methods in each class. We also present publicly available source codes, commonly used datasets, and the most widely utilized evaluation metrics. Finally, we review current trends and suggest future research directions based on the existing challenges.

show abstract

Deep graph similarity learning: a survey

Cited by 60 publications

References 97 publications

SpatialSim: Recognizing Spatial Configurations of Objects With Graph Neural Networks

SpatialSim: Recognizing Spatial Configurations of Objects With Graph Neural Networks

Unsupervised Event Graph Representation and Similarity Learning on Biomedical Literature

Deep Graph Generators: A Survey

Contact Info

Product

Resources

About