Navigating Contextual Complexity with Graph Visualization
Figure 4 – 6 : SysMLv2 Part Interconnection visualization from Tom Sawyer SysMLv2 Viewer .
End users following any prescribed modeling practice , especially in domain-specific modeling contexts , benefit from some standardization at the visualization layer – both in the expectations of specific visualizations , but also in the consistency of automated layouts as models change over time . Constraints on automated layouts can be useful to assert consistency , however , the use of constraints can restrict the effectiveness of automated layout algorithms by forcing them to create less-than-ideal node placement or edge routing .
The examples in Figure 4 – 5 and Figure 4-14 suggest that the answer may lie in providing and utilizing sufficient meta-data to arrive at more consistent automated representations , or a representation that is more true to domain-specific conventions to achieve a higher degree of user acceptance .
4.2 NAVIGATING AND INTERPRETING LARGE GRAPHS
Working with larger graphs presents several obstacles including : where to start analysis , how to focus in on relevant data , and how to communicate key findings . In this example , we explore how to apply visualization and analysis to a social network to understand how agents are working together in a system . In Figure 4-7 , an example social network shows human agents linked if they have worked together .
It is possible to focus into especially relevant parts of the network applying filters which indicate , for example , types of activity , or geospatial locations . It is also possible to apply data-driven analysis which indicates higher importance for certain agents ; or , as in the following figures , combine the findings of human and data-driven analysis .
Journal of Innovation 83