Please enable JavaScript.
Coggle requires JavaScript to display documents.
Visualization of big data (Cluster the data (Criteria (Avoid overlap…
Visualization of big data
Cluster the data
Clustering
Filtering
Criteria
Avoid overlap between clusters
Minimise egde crossing
Minimize edge-cluster crossing
Reduce visual complexity
Matrix representation
Map representation
Edge bundling method
Use 3D
Multiplane algorithm
Divide the large/complex network into clusters
Draw each cluster on a 2D manifold in 3D
Connect the planes with inter-plane edges
Run time: o(n)+2D spring o(n^2)
Scale-free network
Parallel plane layout criteria
Minimize occlusion
Minimize intra-sphere edge crossings
Minimize total inter-plane edge lengths
Integration with analysis
Network analysis
Purpose of interest
Centrality
Cohesive subgroup
Structural roles
Patterns
Centrality analysis
Degree
Distance
Betweenness
Closeness
Eccentricity
K-core analysis
A group analysis method
Identify important dense subgraph from big one
Linear run-time
Network positions and structural equivalence
Block modeling
Structural equivalence
Clusters
Network measures
Degree distribution
Clustering coefficient
Average path length
Density
Connected component
Data mining
Machine learning
6.Integration with interaction
Types of interaction
Filtering
Selecting
Zooming/panning
Collapsing/expanding
Navigating Huge unknown network
Help solving two problems
Visual complexity
Computational complexity
Multi-level approach
Coarsening step
Refinement step