This workflow shows how XPath nodes can be used to extract cluster information from a PMML model. The clusters can be used to do a similarity search for new data, yielding not only the cluster assignment but also a datapoint’s distance to the nearest cluster.