Clustering

Clustering makes it possible to group users and items into a specified number of clusters. For finding cluster, there are three different algorithms that can be used:

1. embedding with item clusters from the trained machine-learning model and users assigned to the most connected cluster,
2. plsa, Probabilistic Latent Semantic Analysis run separately, and
3. modularity, network-based modularity optimizationrun separately.

embedding is the fastest method since it uses the trained machine-learning model.