Graph Processing Module for Edge Bundling and Clutter Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing graph visualization techniques struggle to effectively manage complexity in large data sets, leading to confusing visual renderings, as they fail to adequately bundle and separate edges in a way that minimizes clutter and preserves aesthetic appeal.
Innovation Solution
A graph processing module operates in three phases: edge bundling, edge ordering, and edge separation, using cost functions and candidate-edge graphs to construct bundled paths, order edges, and separate edges, thereby reducing path intersections and improving the visual clarity of the graph.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional graph visualization techniques are used to represent large data sets, then the graph can be displayed with nodes and lines, but the visual rendering becomes complex and confusing
Solution Approach 1:
The patent merges multiple edges that run close together into bundled paths, reducing the visual complexity of the graph while preserving the underlying data relationships. This is achieved by identifying edges with similar trajectories and grouping them into shared bundled paths, thereby reducing clutter and improving visual clarity in large graph data sets
Solution Approach 2:
The patent segments the graph visualization into distinct phases: edge bundling phase, edge ordering phase, and edge separation phase. Each phase processes the graph structure in a specific way to progressively reduce complexity while maintaining information integrity, allowing systematic management of visual complexity
2Shape
If bundling techniques are used to group lines together, then the aesthetic appearance improves, but the existing techniques have room for improvement in managing complexity
Solution Approach 1:
The patent performs preliminary edge bundling before final rendering, identifying and grouping edges into bundled paths in advance. This preliminary organization of edges into aesthetically pleasing bundles before the final visualization step allows the system to achieve both aesthetic appearance and effective complexity management simultaneously
Solution Approach 2:
The patent implements dynamic edge ordering within bundled paths, where the arrangement of edges can be adjusted based on various criteria such as minimizing intersections or optimizing visual flow. This dynamic adjustment capability allows the system to adapt the graph visualization to different viewing contexts while maintaining aesthetic quality
3Loss of information
If edges are bundled together to reduce clutter, then visual clarity improves, but path intersections may increase without proper ordering
Solution Approach 1:
The patent applies partial bundling strategies where not all edges are bundled together, but only those that benefit from bundling without creating excessive intersections. The system selectively bundles edges based on their spatial relationships and the potential for intersection, applying bundling action only where it improves visual clarity without generating harmful path intersections
4Shape
If the graph is processed to achieve desired bundling, then aesthetic appearance improves, but computation time increases due to multiple phases
Solution Approach 1:
The patent segments the graph processing into three distinct phases: edge bundling phase, edge ordering phase, and edge separation phase. This segmentation allows each phase to be optimized independently and processed efficiently, reducing overall computation time while achieving the desired aesthetic appearance through systematic processing of graph edges
Data Source
AI summary
A graph processing module is described for producing a processed graph in three phases. In a bundling phase, the graph processing module selects edges that are used to constitute paths to produce a bundled graph. In an ordering phase, the graph processing module orders paths within the bundled graph to produce an ordered bundled graph. In a separating phase, the graph processing module separates paths in the ordered bundled graph. Among other features, the graph processing module performs the selecting operation by minimizing a cost function; this has the effect of promoting bundling in the bundled graph while discouraging the presence of overly long edges in the bundled graph. The graph processing module can perform the ordering operation by selecting the ordering of paths passing through each edge based on a consideration of ordering constraints within a common subpath associated with the edge.