The discovery of network clustering patterns is an important task in network analysis. A good network layout should reflect the clustering characteristics of the network and allow users to explore the network structure from different levels. For this reason, based on the idea of community division and multi-level layout, this paper proposed a clustering feature hierarchical layout algorithm. First, used seed nodes and personalized PageRank to divide the network. Secondly, it coarsened the network according to the results of the division, and designed the initial layout of the coarsened network. Then it used the node degree information to improve the force-oriented model to complete the refinement. Finally, in order to verify the effectiveness of the algorithm, experiments from the whole to the part were carried out. The experiment shows that the algorithm can generate high-quality layout results within an effective time. Compared with the existing layout algorithms, the proposed algorithm can truly display the network. Clustering features, while taking into account the network microstructure, it can meet the needs of users to explore the network structure from different levels.

[2021-09-19] Accepted Paper /doc/2021.07.0272/v1

[2022-02-05] Printed Article /doc/2021.07.0272/v2