Before you request a clustered result set, you must build the text index with the clustering function enabled. See “Enabling query-by-example, summarization, and clustering”.
The Verity clustering algorithm attempts to group similar rows together, based on the values of the following configuration parameters:
cluster_style
cluster_max
cluster_effort
cluster_order
Use sp_text_cluster to have a query use values that are different from the default values of these configuration parameters. See sp_text_cluster.