Clustering framework

Clustering is a tool for exploring data and detecting patterns that are not apparent to people looking at individual records. Like similarity solutions, a clustering solution quantifies the degree to which records are like one another. The difference is that while similarity search applies to one record at a time, the clustering framework applies to your entire dataset, creating groups of records that mean similar things.

These results are then grouped in a visualization called a treemap, where clusters represent blocks of different records and are sized according to the number of records in each cluster.

Once your clustering solution has been trained, it can be used as an exploratory tool to identify large groups of records that adhere to common patterns. Clustering is very useful for identifying automation opportunities or reoccurring issues.

Creating a clustering solution

Setting up a clustering solution is somewhat more complex than other models...