An Improved Graph Layout Algorithm of Embedded Node Attributes

The traditional graph layout algorithm mainly considers the topology of a graph to generate visually-aesthetic layout results. However without considering the node attribute, the traditional graph layout results can not accurately reflect the impact of the node attribute. Aiming at this problem, this paper proposed an improved graph layout algorithm based on the embedded node attributes. We first define the distance metric between nodes in the node attributes space(including data attribute and topology attribute of the node). Then we define three linear monotonic functions mapping node attribute distances to three force parameters of the force-directed layout algorithm, i.e. gravitation, spring elasticity coefficient and spring original length. According to specific node attributes, we design and calculate the corresponding attribute distance function and the linear monotonic mapping functions to obtain the node-attributes related graph layout results. Finally we apply our algorithm to visualize the Douban movie data and the real political blog data. The experimental results show that the layouts of this algorithm can fully reflect the influence of node attributes, revealing the important node relationships or interesting sub-graph structures associated with node attributes.