sp_text_cluster

Description

(Enhanced version only)

Displays or changes clustering parameters for the active thread.

Syntax

sp_text_cluster server_name, cluster_parameter [, cluster_value]

Parameters

server_name

– is the name of the Full-Text Search engine.

cluster_parameter

– is the name of the clustering parameter. Values are shown in Table A-2.

cluster_value

– is the value you assign to the clustering parameter for the active thread. Values are shown in Table A-2.

Table A-2: Clustering configuration parameters

Values for cluster_parameter

Values for cluster_value

cluster_style

Specifies the type of clustering to use. Valid values are:

  • fixed – generates a fixed number of clusters. The number is set by the cluster_max parameter.

  • coarse – automatically determines the number of clusters to generate, based on fewer, coarse grained clusters.

  • medium – automatically determines the number of clusters to generate, based on medium sized clusters.

  • fine – automatically determines the number of clusters to generate, based on smaller, finer grained clusters.

cluster_max

Specifies the maximum number of clusters to generate when cluster_style is set to fixed. A value of 0 means that the search engine determines the number of clusters to generate.

cluster_effort

Specifies the amount of effort (time) that the search engine should expend on finding a good clustering. Valid values are:

  • effort_default – the search engine spends the default amount of time. You can also use the Verity term “default” if you enclose it in double quotes (“ ”).

  • high – the search engine spends the longest time.

  • medium – the search engine spends less time.

  • low – the search engine spends the least amount of time.

cluster_order

Specifies the order in which to return the rows within the clusters. Valid values are:

  • "0" – indicates rows are returned in order of similarity to the cluster center. This means the first row returned for a cluster is the one that is most prototypical of the rows in the cluster.

  • "1" – indicates that rows are returned in the same relative order in which they were submitted for clustering. For example, if cluster 1 contains the first, third and seventh rows found for the query, they will be returned in that relative order within the cluster.

Examples

Example 1

sp_text_cluster KRAZYKAT, cluster_order, "1"

Changes the cluster_order parameter to 1 for the active thread.

Example 2

sp_text_cluster KRAZYKAT, cluster_style

Displays the current value of the cluster_style parameter.

Usage

Messages

Permissions

Any user can execute sp_text_cluster.

See also

sp_text_configure