Description
Graphs are commonly used to represent relationships within a dataset. They are used across various application domains such as epidemiology (e.g., contact networks), social media (e.g., social networks), and critical infrastructure (e.g., power grid networks). However, in many real-world scenarios, analyzing a single graph in isolation is insufficient. Instead, users must compare multiple networks to uncover meaningful patterns, for instance, when tracking how a time-varying network evolves over time or identifying structural variations across a cohort of subjects.
We are specifically interested in finding effective ways to facilitate the visual comparison of 1-to-Many (comparing a baseline against multiple instances) or Many-to-Many (comparing all-to-all) networks, moving beyond the more straightforward 1-to-1 comparison. This additional layer of complexity requires meaningful visual encoding and interaction techniques to reveal insightful information within the dataset effectively.
Tasks
Review the state of the art in comparative network visualization.
Design and implement at least one new comparison visualization strategy.
Evaluate your proposed approaches.
Requirements
- Basic network analysis knowledge (Python: NetworkX, igraph).
- Knowledge of network/graph visualization techniques.
- Familiarity with interactive visualizations (e.g., D3.js).
- Knowledge in JavaScript
Environment
The project should be implemented as a standalone web-based application (e.g., React).