Topology Preserving Algorithm for Word Clouds with Bounded Constraint

As a classical visualization tool for text analysis, the word cloud has received many attentions. However, the word cloud with bounded constraint may be unstable when words changed. To solve this problem, this paper first proposes a word cloud layout method for certain amount of words. The method uses the Power diagram to divide the space into certain areas. Thereafter, the words are placed into different areas according to their attributes initially. Then this paper adopts five kinds of constraint rules, a relative force between words and a boundary force. These rules and forces help preserve the topology in bounded constraint word clouds. Finally, this paper defines the boundary forces for different shaped boundaries respectively. Furthermore, the experiment results show that the final topology of the word cloud is similar with the original one, when we increase or decrease the size of words or even delete the words in the word cloud.