Summary of "Tutorial 6: Gephi Network Visualization"

This tutorial provides a comprehensive guide on how to use Gephi, an open-source software for network analysis and visualization. It covers data import, network metrics calculation, visualization customization, filtering, workspace management, and exporting results.

Main Ideas and Concepts

Introduction to Gephi
- Open-source software for network analysis and visualization.
- Used in academia, journalism, digital humanities, and social network analysis.
- Supports multiple file formats (gml, graphml, pajek, uci net, dl, CSV, excel).
- Can import social network data from platforms like Facebook and Twitter.
Data Import
- Gephi reads various file formats via File > Open.
- CSV import requires two files:
  - Nodes file: must include a unique id column.
  - Edges file: must include source and target columns, optionally type for directed/undirected edges.
- After import, Gephi shows a summary (number of nodes, edges, etc.).
- Data can be viewed and edited in the Data Laboratory tab (nodes and edges tables).
- Individual values can be edited directly in the table.
Overview Tab & Visualization Setup
- Central workspace for visualizing and manipulating the network.
- Key panels:
  - Appearance: Change node/edge color and size based on categorical or continuous attributes.
  - Layout: Choose and run layout algorithms to arrange nodes (e.g., Fruchterman Reingold, ForceAtlas).
  - Interactive Tools: Select nodes/edges, manually resize or recolor nodes and edges.
  - Node and Edge Pencils: Add new nodes and edges manually.
- Bottom panel allows global adjustments of color, size, and labels.
Filters and Statistics
- Filters:
  - Drag and drop filters to query specific nodes/edges.
  - Types of filters:
    - Attribute-based (equal, range, partition).
    - Edge-based (weight, type, self-loops).
    - Topology-based (components, k-cores, degree).
    - Operator filters to combine multiple filters (intersection, union, complement).
  - Filtered graphs can be saved or exported.
- Statistics:
  - Calculate network metrics by running individual measures:
    - Average degree (with degree distribution).
    - Average weighted degree.
    - Network diameter (average shortest path).
    - Graph density (closeness to complete network).
    - Modularity (community detection).
    - PageRank (node importance).
    - Connected components (strongly and weakly connected).
    - Clustering coefficient.
    - Eigenvector centrality.
  - Computed metrics are added as node attributes and usable for visualization.
Visualization Customization
- Use layout algorithms to improve graph readability.
- Size nodes by attributes (e.g., followers, in-degree).
- Color nodes by categorical attributes (e.g., Modularity class).
- Enable and customize node labels (color, size, font).
- Adjust edge colors and optionally use node colors for edges.
- Use various layouts (Fruchterman Reingold, ForceAtlas, Contraction).
- Zoom, center, and manually adjust node positions using drag tool.
- Use painter tool to manually color nodes.
- Add new nodes and edges with pencils.
- Edit node attributes manually.
- Change background color and toggle between black and white.
- Take screenshots of visualizations in PNG format.
Selection Tools
- Direct selection: select a node and optionally its neighbors.
- Rectangular selection: select nodes within a rectangular area, with or without neighbors.
- Auto-select neighbor option can be toggled on/off.
Filters and Workspace Management
- Apply filters to isolate subsets of the network.
- Export filtered graphs to new workspaces for focused analysis.
- Each workspace is independent, allowing separate filtering, visualization, and layout.
- Switch between workspaces to compare subsets and original graph.
Preview Tab
- Advanced visualization adjustments.
- Options to toggle curved edges, rescale edge thickness.
- Show/hide node labels and adjust font size.
- Export final visualization in PDF, PNG, or SVG formats.
Navigation Between Data Laboratory and Overview
- Select nodes in visualization and highlight corresponding rows in Data Laboratory.
- Select rows in Data Laboratory and highlight nodes in visualization.
- Facilitates detailed examination of nodes.

Detailed Methodology / Instructions

Importing Data
- File > Open > Select file (gml, graphml, CSV, etc.).
- For CSV:
  - Prepare two files: nodes (with id) and edges (with source, target, optional type).
  - Use Import Spreadsheet button if needed.
Viewing