DeepGD: A Deep Learning Framework for Graph Drawing Using GNN

TL;DR

DeepGD is a novel deep learning framework using GNNs that can generate aesthetically pleasing graph layouts for any graph after training, balancing multiple aesthetic criteria adaptively.

Contribution

It introduces a generalizable GNN-based graph drawing method with adaptive training strategies to optimize multiple aesthetics simultaneously.

Findings

Effective at drawing arbitrary graphs

Capable of balancing multiple aesthetic criteria

Flexible and generalizable approach

Abstract

In the past decades, many graph drawing techniques have been proposed for generating aesthetically pleasing graph layouts. However, it remains a challenging task since different layout methods tend to highlight different characteristics of the graphs. Recently, studies on deep learning based graph drawing algorithm have emerged but they are often not generalizable to arbitrary graphs without re-training. In this paper, we propose a Convolutional Graph Neural Network based deep learning framework, DeepGD, which can draw arbitrary graphs once trained. It attempts to generate layouts by compromising among multiple pre-specified aesthetics considering a good graph layout usually complies with multiple aesthetics simultaneously. In order to balance the trade-off, we propose two adaptive training strategies which adjust the weight factor of each aesthetic dynamically during training. The quantitative and qualitative assessment of DeepGD demonstrates that it is capable of drawing arbitrary graphs effectively, while being flexible at accommodating different aesthetic criteria.