Understanding the power of communities
Previously, in Chapter 9, we looked at finding similar customers using vector similarity and how to provide recommendations for a customer. Let’s revisit Figure 9.10 and Figure 9.11. Figure 9.10 displays the purchase history of customers similar to a particular customer. Figure 9.11 shows the recommendations for a customer based on the purchases of similar customers. The purchase history and customer recommendations are a result of the Cypher queries we worked on in the Fine-tuning your recommendations section to understand vector similarity usage.
In this section, we will take a deeper look at communities and see why they might be better than leveraging simple vector similarity to find similar customers.
Note
The following Cyphers are related to the database shared in the Technical
requirements section.
From the Cypher we ran in the last section, Detecting communities with the Louvain algorithm...
