Lexical navigation: visually prompted query expansion and refinement
DL '97 Proceedings of the second ACM international conference on Digital libraries
Web-based animation of data structures using JAWAA
SIGCSE '98 Proceedings of the twenty-ninth SIGCSE technical symposium on Computer science education
Empirical Evaluation of Aesthetics-based Graph Layout
Empirical Software Engineering
JIGGLE: Java Interactive Graph Layout Environment
GD '98 Proceedings of the 6th International Symposium on Graph Drawing
Human-centered visualization environments
Human-centered visualization environments
| Hi-index | 0.00 |
FTP: CMU-CS-94-161.ps Although there is extensive research on drawing graphs, none of the published methods are satisfactory for drawing general undirected graphs. Generating drawings which are optimal with respect to several aesthetic criteria is known to be NP-hard, so all published approaches to the problem have used heuristics. These heuristics are too slow to be practical for graphs of moderate size, and they do not produce consistently good drawings for general graphs. Moreover, they rely on general optimization methods, because problem-specific methods require a deeper theoretical understanding of the graph drawing problem. This paper presents an algorithm to generate two-dimensional drawings of undirected graphs. The algorithm uses a combination of heuristics to obtain drawings which are near-optimal with respect to an aesthetic cost function. The algorithm is incremental in nature, but preprocesses the graph to determine an order for node placement. The algorithm uses a local optimization strategy that effectively manages the trade-off between speed and output quality. Finally, the algorithm uses a variety of techniques to speed up computation of the aesthetic cost function. The paper discusses this algorithm in the context of previous work and open problems. The algorithm is compared with the "force-directed" algorithm of Fruchterman and Reingold and the simulated annealing algorithm of Davidson and Harel in terms of output quality. Finally, the paper considers what work is necessary to create a truly effective algorithm for drawing undirected graphs.