This workshop will provide a comprehensive introduction to the problem of Community Detection (CD), i.e. dividing a network into groups of nodes, along with materials and code for its implementation. It will cover the main families of algorithms and the major challenges in implementation, as well as means to compare and evaluate solutions. The workshop will mostly focus on the standard problem of non-overlapping partitions on undirected, unweighted networks, although other problem and network variations will be briefly reviewed at the end. While the code and examples will be provided in python within an interactive online platform, so familiarity with the language will not be required and there is no setup needed previous to the workshop.
Overview of content:
- Introduction to the problem and applications
- Overview of the main families of problem definitions and the most common algorithms, and their implementations
- Evaluation of CD solutions
- Comparing and combining multiple solutions
- Quick overview of variations: overlapping CD, multiplex, temporal