4-4
Cisco Personal Assistant 1.4 Installation and Administration Guide
OL-4590-03
Chapter 4 Configuring Personal Assistant
Configuring Speech Recognition
Step 4 Click Session > Logout in the menu bar to log out of the system when you are finished configuring
Personal Assistant.
The system automatically logs you out after 30 minutes of inactivity.
Configuring Speech Recognition
You must configure speech services to create speech server clusters. This allows users to talk to
Personal Assistant over the phone, for example, dialing parties by name rather than by phone number.
See the “To Configure Speech Recognition” procedure on page 4-4. If you are not using speech
recognition with Personal Assistant, you can skip this configuration.
You can also fine-tune speech recognition if users are having problems with the speech recognition
interface. See the “To Fine-Tune Speech Recognition” procedure on page 4-5. Problems that you can
address include:
• Users being asked too often to confirm their commands.
• Users being transferred to the operator too often because:
–
There are a large number of similarly named people in your corporate directory.
–
Errors occur in Personal Assistant understanding spoken commands.
–
Users ask for help from Personal Assistant too often during a session.
Make sure you have the license document included in the Personal Assistant package.
Personal Assistant uses defaults for all speech recognition parameters, and these should work well with
most installations. Before fine-tuning the speech recognition parameters, confirm that any reported
problems are widespread. If changes are necessary, trial and error is the only way to determine effective
settings for these parameters for your specific needs.
To Configure Speech Recognition
Step 1 In the Personal Assistant Administration, select System > Speech Services. The Speech Services
Configuration page opens.
Step 2 If you want Personal Assistant to automatically refresh cached user information from the corporate
directory on a daily schedule, check the Daily Automatic Refresh check box, and set the time by using
the Refresh Schedule hour and minute lists. This ensures that the corporate directory information stays
up-to-date, and that the speech recognition software has the data required to understand spoken
commands.
Note Depending on the size of your corporate directory, the number of locales installed, and the
number of directory hierarchies configured, it can take a significant amount of time to fetch this
information, compile it, and publish it to the speech servers. Pick a time when directory updates
are complete for the day, and there is reduced user access to Personal Assistant and the directory.
Step 3 If you want Personal Assistant to send you the refresh status by e-mail when a refresh has completed,
check the Send Refresh Status check box, and enter your e-mail address in the Administrator E-mail
Address field.
Step 4 In the License Key field, enter your speech recognition license key.