Real-world graphs are neither regular nor fully random grids. Their edge densities are not homogeneous, so we end up finding some interesting patterns. The fact that some nodes can have more connections than others is exploited in centrality algorithms to assess node importance (see Chapter 6, Node Importance). In this chapter, we will discover a new type of algorithm whose goal is to identify groups of nodes highly connected to each other and form a community or cluster. Several of these community detection algorithms are already implemented in the GDS: components algorithms, Label Propagation algorithms, and Louvain algorithms. This chapter is our opportunity to build a graph representation of ...
Get Hands-On Graph Analytics with Neo4j now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
