Computer Engineering and Applications ›› 2023, Vol. 59 ›› Issue (16): 1-15.DOI: 10.3778/j.issn.1002-8331.2212-0068

• Research Hotspots and Reviews • Previous Articles     Next Articles

Review of Latest Research for Layout Methods of Graph Visualization

YANG Zhuo, XIE Yaqi, CHEN Yi, ZHAN Yinwei   

  1. 1.School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, China
    2.School of Computer Science and Engineering, Beijing Technology and Business University, Beijing 100048, China
  • Online:2023-08-15 Published:2023-08-15

图可视化布局方法最新研究进展综述

杨卓,谢雅淇,陈谊,战荫伟   

  1. 1.广东工业大学 计算机学院,广州 510006
    2.北京工商大学 计算机学院,北京 100048

Abstract: Graph visualization is an intuitive representation of graph data. With the wide application of graph processing, a decent graph visualization can lead to profound and efficient understanding of graph data. However, as the growth of graph data volume, the layout of graph visualization encounters the challenges of long computation time, difficulties in discovering important structures and relations in graph, and visual clutter caused by node overlapping and complex edge crossing. Therefore, how to lay out large-scale graph data fast, how to enhance the exploration of the important structures and relations in graph, and how to generate the aesthetic layout of graph data become urgent problems to be solved. In recent decades, various optimization methods based on stress models and aesthetic criteria are proposed to address these problems. In addition, machine learning methods such as graph mining, graph embedding and graph neural network provide novel alternative solutions from the perspective of graph data features for the layout of graph visualization. In comparison, machine learning methods exceed in the efficiency and performance of the layout of graph data. This paper briefly summaries the latest research on the layout of graph visualization from four perspectives, which are force-directed algorithm, aesthetic constraint based methods, graph mining algorithm and machine learning methods. Last but not least, the future development of layout methods for graph visualization is discussed.

Key words: layout of graph visualization, node-link graph, force-directed algorithm, graph mining algorithm, aesthetic criteria, machine learning

摘要: 图可视化是图数据的直观表示,随着图数据的广泛应用,合适的图可视化能够使用户对图数据的理解更加深入和高效。但随着图数据量级的增长,图可视化布局面临着计算时间长,难以发现图的重要结构和关系,以及节点遮挡和复杂的边交叉所产生的视觉杂乱等挑战。因此,如何快速对大规模图数据进行布局,如何强化对图中重要的结构和关系的探索,以及如何生成美观的图可视化布局成为亟需解决的问题。近年来,许多基于力学模型和美学评价标准的优化方法被提出来解决上述问题。另外,图挖掘、图嵌入、图神经网络等机器学习方法从图数据特点的角度,为解决图可视化的布局问题提供了新思路,相比之下,机器学习方法在布局效率和效果上表现出一定的优越性。主要从力导向算法、基于美学约束的布局方法、图挖掘技术和机器学习方法这四方面对图可视化布局的最新研究进展进行了阐述,最后对图可视化布局方法的未来发展进行了展望。

关键词: 图可视化布局, 节点-链接图, 力导向算法, 图挖掘算法, 美学评价标准, 机器学习