Clustering of mutual links. What options?

Question

PHP

There are users, for example, about 100k, who like each other's photos. The task is to identify user communities whose likes are distributed only within a small group of 5 to 100 people (mutual likes are immediately cut off, we are talking about longer chains within this group).
What are the algorithms for clustering such things? To detect such anomalies.

Answer 1

Search for strongly connected components in a graph.
https://ru.wikipedia.org/wiki/Component_strong_s...