Abstract
Abstract Hairball buster (HB) (also called node-neighbor centrality or NNC) is an approach to graph analytic triage that uses simple calculations and visualization to quickly understand and compare graphs. Rather than displaying highly interconnected graphs as ‘hairballs’ that are difficult to understand, HB provides a simple standard visual representation of a graph and its metrics, combining a monotonically decreasing curve of node metrics with indicators of each node’s neighbors’ metrics. The HB visual is canonical, in the sense that it provides a standard output for each node-link graph. It helps analysts quickly identify areas for further investigation, and also allows for easy comparison between graphs of different data sets. The calculations required for creating an HB display is order M plus N log N, where N is the number of nodes and M is the number of edges. This paper includes examples of the HB approach applied to four real-world data sets. It also compares HB to similar visual approaches such as degree histograms, adjacency matrices, blockmodeling, and force-based layout techniques. HB presents greater information density than other algorithms at lower or equal calculation cost, efficiently presenting information in a single display that is not available in any other single display.
Highlights
Hairball buster (HB) is an approach to graph analytic triage that uses simple calculations and visualization to quickly understand and compare graphs
The purpose of this paper is to describe a new method for analyzing relationships among nodes in a graph using a canonical representation that enables comparison between different graphs
HB computes a centrality measure for a node and its neighbors, and presents this computation in an efficient, standardized visual form that scales to very large graphs
Summary
Hairball buster (HB) ( called node-neighbor centrality or NNC) is an approach to graph analytic triage that uses simple calculations and visualization to quickly understand and compare graphs. While displaying the full data set provides the information described above for the first 28 nodes, it does not definitively indicate whether the highest degree nodes are fully connected, or whether they belong to separate clusters due to visual occlusion.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.