Community Discovery algorithms¶
CDlib collects implementations of several Community Discovery algorithms.
To maintain the library organization as clean and resilient as possible the approaches are grouped following a simple, two level, rationale:
- The first distinction is made on the object clustered, thus separating Node Clustering and Edge Clustering algorithms;
- The second distinction is made on the specific kind of partition each one of them generates: Crisp, Overlapping or Fuzzy.
This documentation follows the same rationale.
CDlib implements also time-aware algorithms (often referred as Dynamic Community Discovery approaches).