Install the Classification Precision and Recall Server
- Last Updated: May 13, 2026
- 2 minute read
- Semaphore
- Documentation
Note: Before installing, see Requirements software prerequisites or required technical specifications for Semaphore.
The Classification Precision and Recall Server is a tool that allows users to assess the quality of their classification. When the Classification Precision and Recall Server is enabled, you send a set of control documents along with a spreadsheet containing the expected the results to the Classification Service. The Classification Precision and Recall Server returns a spreadsheet that provides detailed information about the quality of the classification, including differences between the expected and returned results. For general information about the Classification Precision and Recall Server, refer to Semaphore Classification Precision and Recall Server.
If required, on the target Semaphore server do the following to install Classification Precision and Recall on Windows:
-
On the target server, run the following installer program executable as an administrator:
Semaphore-<Version>-CLS-PandR-setup.exe -
The License Agreement window appears. Make sure that you read and understand the license agreement. To continue with the installation, select the I accept the agreement; then, click Next.
-
In the Select Destination Location window, specify the location to install the Classification Server. Then click Next.
-
In the Java Home Configuration window, specify the location of the JVM (
jvm.dll) to be used by Semaphore. See Installing third-party software for supported Java versions. -
The Reconciliation Server configuration window opens. Specify the port on which the Studio server listens. Note that the specified port must be free. Typically, the default value is fine. Click Next to proceed.
-
The Select Start Menu Folder window appears. Provide the location in the Start Menu to add Semaphore shortcuts. Click Next.
-
Click Install to begin the installation.
After all setup dialogs have been completed click Finish to close the last dialogue. The service will be available on
http://<machine_name>:<port>
For example:
http://localhost:5091
See also: