UML class diagrams are helpful for understanding the structure of a software system. Algorithms and tools have been developed to generate UML class diagrams automatically for program understanding purposes. However, many tools often ignore perceptual factors in the layout of these diagrams. Therefore, users still have to spend much time and effort rearranging boxes and lines to make the diagram understandable. This paper presents key criteria and guidelines for the effective layout of UML class diagrams from the perspective of perceptual theories. Two UML tools have been analyzed and evaluated to illustrate how the criteria can be applied to improve the readability of class diagrams.

Graphs are typically visualized as node-link diagrams. Although there is a fair amount of research focusing on crossing minimization to improve readability, little attention has been paid on how to handle crossings when they are an essential part of the final visualizations. This requires us to understand how people read graphs and how crossings affect reading performance. As an initial step to this end, a preliminary eye tracking experiment was conducted. The specific purpose of this experiment was to test the effects of crossing angles and geometric-path tendency on eye movements and performance. Sixteen subjects performed both path search and node locating tasks with six drawings. The results showed that small angles can slow down and trigger extra eye movements, causing delays for path search tasks, whereas crossings have little impact on node locating tasks. Geometric-path tendency indicates that a path between two nodes can become harder to follow when many branches of the path go toward the target node. The insights obtained are discussed with a view to further confirmation in future work.

Force-directed layout is typically used to create organic-looking, straight-edge drawings of large graphs while combinatorial techniques are generally preferred for high-quality layout of small to medium sized graphs. In this paper we integrate edge-routing techniques into a forcedirected layout method based on constrained stress majorisation. Our basic procedure takes an initial layout for the graph, including poly-line paths for the edges, and improves this layout by moving the nodes to reduce stress and moving edge bend points to straighten the edges and reduce their overall length. Separation constraints between nodes and edge bend points are used to ensure that node labels do not overlap edges or other nodes and that no additional edge crossings are introduced.

Abstract—Exploring communities is an important task in social network analysis. Such communities are currently identified using clustering methods to group actors. This approach often leads to actors belonging to one and only one cluster, whereas in real life a person can belong to several communities. As a solution we propose duplicating actors in social networks and discuss potential impact of such a move. Several visual duplication designs are discussed and a controlled experiment comparing network visualization with and without duplication is performed, using 6 tasks that are important for graph readability and visual interpretation of social networks. We show that in our experiment, duplications significantly improve community-related tasks but sometimes interfere with other graph readability tasks. Finally, we propose a set of guidelines for deciding when to duplicate actors and choosing candidates for duplication, and alternative ways to render them in social network representations. Index Terms—Clustering, Graph Visualization, Node Duplications, Social Networks. 1

Reprinted with permission.

Abstract — Interactive visualization requires the translation of data into a screen space of limited resolution. While currently ignored by most visualization models, this translation entails a loss of information and the introduction of a number of artifacts that can be useful, (e.g., aggregation, structures) or distracting (e.g., over-plotting, clutter) for the analysis. This phenomenon is observed in parallel coordinates, where overlapping lines between adjacent axes form distinct patterns, representing the relation between variables they connect. However, even for a small number of dimensions, the challenge is to effectively convey the relationships for all combinations of dimensions. The size of the dataset and a large number of dimensions only add to the complexity of this problem. To address these issues, we propose Pargnostics, parallel coordinates diagnostics, a model based on screen-space metrics that quantify the different visual structures. Pargnostics metrics are calculated for pairs of axes and take into account the resolution of the display as well as potential axis inversions. Metrics include the number of line crossings, crossing angles, convergence, overplotting, etc. To construct a visualization view, the user can pick from a ranked display showing pairs of coordinate axes and the structures between them, or examine all possible combinations of axes at once in a matrix display. Picking the best axes layout is an NP-complete problem in general, but we provide a way of automatically optimizing the display according to the user’s preferences based on our metrics and model. Index Terms—Parallel coordinates, metrics, display optimization, visualization models. 1

A graph G is an α angle crossing (αAC) graph if every pair of crossing edges in G intersect at an angle of at least α. The concept of right angle crossing (RAC) graphs (α = π/2) was recently introduced by Didimo et al. [6]. It was shown that any RAC graph with n vertices has at most 4n −10 edges and that there are infinitely many values of n for which there exists a RAC graph with n vertices and 4n − 10 edges. In this paper, we give upper and lower bounds for the number of edges in αAC graphs for all 0 < α < π/2. 1

Effectively visualizing complex node-link graphs which depict relationships among data nodes is a challenging task due to the clutter and occlusion resulting from an excessive amount of edges. In this paper, we propose a novel energy-based hierarchical edge clustering method for node-link graphs. Taking into the consideration of the graph topology, our method first samples graph edges into segments using Delaunay triangulation to generate the control points, which are then hierarchically clustered by energy-based optimization. The edges are grouped according to their positions and directions to improve comprehensibility through abstraction and thus reduce visual clutter. The experimental results demonstrate the effectiveness of our proposed method in clustering edges and providing good high level abstractions of complex graphs.

Cutting-edge biological and bioinformatics research seeks a systems perspective through the analysis of multiple types of high-throughput and other experimental data for the same sample. Systemslevel analysis requires the integration and fusion of such data, typically through advanced statistics and mathematics. Visualization is a complementary computational approach that supports integration and analysis of complex data or its derivatives. We present a bioinformatics visualization prototype, Juxter, which depicts categorical information derived from or assigned to these diverse data for the purpose of comparing patterns across categorizations. The visualization allows users to easily discern correlated and anomalous patterns in the data. These patterns, which might not be detected automatically by algorithms, may reveal valuable information leading to insight and discovery. We describe the visualization and interaction capabilities and demonstrate its utility in a new field, metagenomics, which combines molecular biology and genetics to identify and characterize genetic material from multi-species microbial samples.

Fig. 1. LastHistory: Visualizing personal music listening histories, photo and calendar streams for analysis and reminiscing. Abstract—The choices we take when listening to music are expressions of our personal taste and character. Storing and accessing our listening histories is trivial due to services like Last.fm, but learning from them and understanding them is not. Existing solutions operate at a very abstract level and only produce statistics. By applying techniques from information visualization to this problem, we were able to provide average people with a detailed and powerful tool for accessing their own musical past. LastHistory is an interactive visualization for displaying music listening histories, along with contextual information from personal photos and calendar entries. Its two main user tasks are (1) analysis, with an emphasis on temporal patterns and hypotheses related to musical genre and sequences, and (2) reminiscing, where listening histories and context represent part of one’s past. In this design study paper we give an overview of the field of music listening histories and explain their unique characteristics as a type of personal data. We then describe the design rationale, data and view transformations of LastHistory and present the results from both a lab- and a large-scale online study. We also put listening histories in contrast to other lifelogging data. The resonant and enthusiastic feedback that we received from average users shows a need for making their personal data accessible. We hope to stimulate such developments through this research.