Speeding-Up Forced-directed Algorithm by Clustering and Its Implementation

• Main Authors: Ho, Chia Han, 何佳翰
• Other Authors: Lee, Che Rung
• Format: Others
• Language: en_US
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/93325593382657575029

碩士 === 國立清華大學 === 資訊工程學系 === 104 === Graph drawing can simplify complex information of graphs through visualization. One of commonly used graph drawing methods is Force-directed algorithm (FDA) because of its easy of implementation and good quality of uniform edge length, uniform vertex distribution, etc. However, classical FDA has a high running time when the number of vertices is large. In this thesis, we present a modified FDA algorithm which can reduce the running time of FDA. The major idea is to partition the graph into clusters, to run FDA of each clusters in parallel, and to merge the results in the end. We evaluated our method on several data sets, and the experimental results show the proposed method can efficiently reduce the running time without scarifying the quality of the classical FDA.