In this chapter, we talked a lot about ways to measure the similarity between nodes, either on a global scale by grouping nodes into communities or with a more local similarity assessment, for example, using the Jaccard similarity metric. Several algorithms were studied – the weakly and strongly connected components, the Label Propagation algorithm, and the Louvain algorithms. We also used a feature offered by the GDS that allows us to write the results of an algorithm into Neo4j for future use. We also used two new tools to visualize a graph and the results of the graph algorithms implemented in the GDS: neovis.js, which is used to embed a Neo4j graph visualization into an HTML page, and NEuler, which is the Graph Algorithms Playground, from which you can run a graph algorithm without writing code.
Our exploration of the algorithms implemented in the GDS (1.0) is now finished. In the next chapters, we will learn how to use graphs and these algorithms in a machine learning...
