Machine Learning on Graphs: A Model and Comprehensive Taxonomy

By Ines Chami, Sami Abu-El-Haija, Bryan Perozzi, Christopher R'e, Kevin P. Murphy. Journal of machine learning research, 2020

347 citations recorded by the source corpus.

There has been a surge of recent interest in learning representations for graph-structured data. Graph representation learning methods have generally fallen into three main categories, based on the availability of labeled data. The first, network embedding (such as shallow graph embedding or graph auto-encoders), focuses on learning unsupervised representations of relational structure. The second, graph regularized neural networks, leverages graphs to augment neural network losses with a regularization objective for semi-supervised learning. The third, graph neural networks, aims to learn differentiable functions over discrete topologies with arbitrary structure. However, despite the popularity of these areas there has been surprisingly little work on unifying the three paradigms. Here, we aim to bridge the gap between graph neural networks, network embedding and graph regularization models. We propose a comprehensive taxonomy of representation learning methods for graph-structured data, aiming to unify several disparate bodies of work. Specifically, we propose a Graph Encoder Decoder Model (GRAPHEDM), which generalizes popular algorithms for semi-supervised learning on graphs (e.g. GraphSage, Graph Convolutional Networks, Graph Attention Networks), and unsupervised learning of graph representations (e.g. DeepWalk, node2vec, etc) into a single consistent approach. To illustrate the generality of this approach, we fit over thirty existing methods into this framework. We believe that this unifying view both provides a solid foundation for understanding the intuition behind these methods, and enables future research in the area.

Subfields: Artificial Intelligence, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/83e89d56d0d0e1dfd8b52213e6cc2e191aaaf34b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistical and Nonlinear Physics

Citations in the corpus

Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs (2020, background): “Besides the models mentioned above, there are various comprehensive reviews describing previously proposed architectures [42, 5, 41].”
Utilizing graph machine learning within drug discovery and development (2021, background): “Most machine learning methods that operate on graphs can be decomposed into two parts: a general-purpose encoder and a taskspecific decoder [62].”
Graph Clustering with Graph Neural Networks (2020, background): “For a complete introductions to the vast topic we refer interested readers to detailed surveys [6, 8, 25].”
TUDataset: A collection of benchmark datasets for learning with graphs (2020, background): “A survey of recent advancements in GNN techniques can be found, e.g., in (Chami et al., 2020; Wu et al., 2019; Zhou et al., 2018).”
Artificial Neural Networks and Machine Learning – ICANN 2021: 30th International Conference on Artificial Neural Networks, Bratislava, Slovakia, September 14–17, 2021, Proceedings, Part III (2021, background): “The prime example is perhaps constituted by data naturally represented as graphs, which poses significant challenges to Euclidean-based learning [44,5].”
Scaling Graph Neural Networks with Approximate PageRank (2020, neutral): “[22] or other similar frameworks [6, 12, 50].”
Combinatorial optimization and reasoning with graph neural networks (2021, background): “Some of the most recent ones include [19, 141, 153].”
Adversarial Graph Augmentation to Improve Graph Contrastive Learning (2021, methodology): “1, the graph representation is obtained by pooling the ﬁnal set of node representations as, For design choices regarding aggregation, update and pooling functions we refer the reader to [3,7,8].”