System Administration Guide for Cisco Unity Release 8.x (With Microsoft Exchange) - Managing Voice-Recognition Features in Cisco Unity 8.x [Cisco Unity]

Table Of Contents

Managing Voice-Recognition Features in Cisco Unity 8.x

Access to Voice Recognition Features in Cisco Unity 8.x

Confirming Voice Commands Based on Recognition Confidence Scores in Cisco Unity 8.x

Voice Addressing in Cisco Unity 8.x

Managing the Alternate Names List

Updating the Name Grammar Files

Voice-Recognition Considerations in Cisco Unity 8.x

General Considerations for Subscribers

Custom Key Map Conversations

Improving Voice Recognition Performance

Disabling Voice Recognition for All Subscribers in Cisco Unity 8.x

Managing Voice-Recognition Features in Cisco Unity 8.x

Voice-recognition features allow subscribers to interact with Cisco Unity by speaking rather than by using the phone keypad. The features are licensed by number of concurrent sessions, and require installation and configuration of voice-recognition software on a separate and dedicated server. Voice recognition is not available to outside callers.

You configure voice recognition settings on the System > Voice Recognition pages in the Cisco Unity Administrator.

For information on installing and configuring a separate voice-recognition server, see the applicable Installation Guide for Cisco Unity, at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/prod_installation_guides_list.html.

See the following sections in this chapter for additional details about administering voice-recognition features:

•Access to Voice Recognition Features in Cisco Unity 8.x

•Confirming Voice Commands Based on Recognition Confidence Scores in Cisco Unity 8.x

•Voice Addressing in Cisco Unity 8.x

•Voice-Recognition Considerations in Cisco Unity 8.x

•Disabling Voice Recognition for All Subscribers in Cisco Unity 8.x

Access to Voice Recognition Features in Cisco Unity 8.x

You enable access to the voice-recognition features on the applicable Conversation page for a subscriber template or an individual subscriber in the Cisco Unity Administrator, or by using the Bulk Edit utility.

Subscribers who have access to voice-recognition features can turn voice recognition on or off for themselves by phone or by using the Cisco Unity Assistant. Subscribers who have voice recognition (also known as the Press-or-Say phone input style) turned on can do the following:

•Say their voicemail passwords when logging on to the Cisco Unity telephone user interface.

•Press the phone key, say the number, or say the name of a menu option. For example, in a menu where three corresponds to a delete action, subscribers can press three, say "Three," or say "Delete."

•Say dates and times when taking actions that require them, such as entering an end date and time for the alternate greeting.

•Say the names of recipients instead of spelling them with the phone keypad when addressing messages, adding names to private lists, and using the Message Locator feature.

Tip To improve the accuracy of name recognition, you can enter alternate spellings for the names of individual subscribers. Alternate spellings are entered on the Alternate Name page for individual subscribers in the Cisco Unity Administrator. See the "Voice Addressing in Cisco Unity 8.x" section for more information.

•Say "Hang up" at any menu to disconnect the call.

Do the following "To Enable Access to Voice-Recognition Features" procedure to enable voice-recognition access for subscribers, and to turn on voice recognition, if desired. (To make the changes for a group of subscribers, you can use the Bulk Edit tool available in Tools Depot.)

Note The voice recognition features require an additional server running voice-recognition software. For information on installing and configuring a separate voice-recognition server, see the applicable Installation Guide for Cisco Unity, at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/prod_installation_guides_list.html.

To Enable Access to Voice-Recognition Features

Step 1 In the Cisco Unity Administrator, go to the applicable page:

•To modify the template that you will use to create subscriber accounts, go to any Subscribers > Subscriber Template page, and find the template that you want to modify. Then browse to the Conversation page.

•To modify an existing subscriber account, go to any Subscribers > Subscribers page and find the applicable subscriber. Then browse to the Conversation page.

Step 2 Under Phone Menu Options, check the Allow Access to Voice-Recognition Features check box.

Step 3 Optionally, to turn on voice recognition, check the Use Press-or-Say Phone Input Style (Includes Voice Addressing) check box.

Step 4 Click the Save icon.

Confirming Voice Commands Based on Recognition Confidence Scores in Cisco Unity 8.x

In certain situations, subscribers may want to be prompted to confirm an important action if the confidence score on their utterance falls below a certain level. For example, if the voice-recognition server matches an utterance to the "delete" action with a confidence level of 60%, Cisco Unity can prompt the user to confirm the deletion before proceeding.

The Threshold to Trigger Speech Verification value allows you to specify the confidence level that triggers the confirmation. The setting applies to voice command actions for deleting messages; addressing a reply to all recipients; sending, replying to, or forwarding a message; or calling the sender of a message (live reply). The default confidence level is 85.

To Configure the Threshold to Trigger Speech Verification

Step 1 In the Cisco Unity Administrator, go to the System Settings > Voice Recognition > Settings page.

Step 2 Under Voice Recognition Server, enter a value for the Threshold to Trigger Speech Verification setting. The range of accepted values is 0 (never prompt for confirmation) to 100 (always prompt for confirmation).

Step 3 Click the Save icon.

Step 4 If the Cisco Unity system is configured for failover, repeat this procedure to configure the threshold for the secondary server. Voice recognition settings are not replicated between the primary and secondary servers.

Voice Addressing in Cisco Unity 8.x

Cisco Unity recognizes spoken subscriber names, so subscribers do not have to spell them by saying letters. They can say subscriber names while addressing messages, adding subscribers to private lists, using Message Locator, or addressing public distribution lists.

See the following sections:

•Managing the Alternate Names List

•Updating the Name Grammar Files

For additional information on improving name recognition, see the "Voice-Recognition Considerations in Cisco Unity 8.x" section.

Managing the Alternate Names List

Alternate names are different versions of a subscriber name than what is listed in the corporate directory. Cisco Unity considers these names when a subscriber uses voice recognition to address messages, add subscribers to private lists, use Message Locator, or address public distribution lists.

You create alternate names for individual subscribers, and cannot add or edit alternate names from the subscriber template. Also, it is important to note that any alternate names you create are stored locally and are not propagated to other Cisco Unity servers. Any alternate names that you add or modify are not active until there has been a grammar rebuild. See the "Updating the Name Grammar Files" section for more information on rebuilding the name grammar files.

When creating alternate names, you might want to add common variants of a name that subscribers might use. For example, you might want to add Bill, Billy, and Will for the name William. Or, you could use this feature to add another version of an uncommon name, unusual nicknames, or maiden names. For example, if a subscriber asked Cisco Unity to address "Mary Jameson," which was the maiden name of Mary Brown, Cisco Unity could reference this information and add the correct subscriber to the message. You could also use alternate names to add phonetic spellings of hard-to-pronounce names. For example, you could add "Goolay" as an alternate name for the last name "Goulet."

Note that name recognition accuracy degrades as the number and variability of names increase. In addition, name recognition tends to be less accurate for non-English names. For this reason, we recommend that you include only the most commonly-known alternate names for subscribers.

See the following procedures:

•To Add Alternate Names

•To Edit Alternate Names

Note that you cannot add or edit alternate names on a subscriber template.

To Add Alternate Names

Step 1 In the Cisco Unity Administrator, go to any Subscribers > Subscribers page and find the applicable subscriber. Then browse to the Alternate Names page.

Step 2 Under Add, in the Alternate Names fields, enter the alternate names.

Step 3 Click Add.

Step 4 Repeat Step 2 and Step 3 until all alternate names have been added.

To Edit Alternate Names

Step 1 In the Cisco Unity Administrator, go to any Subscribers > Subscribers page and find the account that you want to change. Then browse to the Alternate Names page.

Step 2 Do any of the following:

•In the Alternate Names fields, enter changes to the already-existing alternate names.

•If you want to delete an alternate name, check the check box next to the name.

Step 3 Click the Save icon.

Updating the Name Grammar Files

The name grammar files should be updated immediately after Cisco Unity installation and configuration, and updated routinely anytime Cisco Unity subscribers are created, added or imported to the Cisco Unity directory. The names of new Cisco Unity subscribers and their alternate names will not be immediately available to the voice-recognition system unless the grammar files are updated.

By default, the name grammar files are scheduled to update automatically at 1:00 a.m. daily. To make changes available immediately or at a different scheduled time, do one of the following procedures:

•To Schedule When Cisco Unity Automatically Rebuilds the Name Grammar Files

•To Manually Rebuild the Name Grammar Files

Note Depending on the size of the directory, name grammars may take a long time to build and may affect system performance. We recommend that you schedule the automatic rebuild to occur at off-peak times. By default, the automatic rebuild starts at 1:00 a.m. Manual updates are recommended only when troubleshooting or during the initial Cisco Unity configuration.

To Schedule When Cisco Unity Automatically Rebuilds the Name Grammar Files

Step 1 In the Cisco Unity Administrator, go to the System > Voice Recognition > Settings page.

Step 2 Under Grammar Engine, in the Automatic Rebuild Time fields, enter the new time.

By default, the automatic rebuild starts at 1:00 a.m. We recommend that you schedule the automatic rebuild to occur at off-peak times.

Step 3 Click the Save icon.

Step 4 If the Cisco Unity system is configured for failover, repeat this procedure to configure the threshold for the secondary server. Voice recognition settings are not replicated between the primary and secondary servers.

To Manually Rebuild the Name Grammar Files

Depending on the size of the directory, name grammars may take a long time to build and may affect system performance. Manual updates are recommended only when troubleshooting or during the initial Cisco Unity configuration.

Step 1 In the Cisco Unity Administrator, go to the System > Voice Recognition > Settings page.

Step 2 Under Grammar Engine, click Start Manual Update.

Step 3 If the Cisco Unity system is configured for failover, repeat this procedure to update the name grammar files on the secondary server.

Voice-Recognition Considerations in Cisco Unity 8.x

See the following sections:

•General Considerations for Subscribers

•Custom Key Map Conversations

•Improving Voice Recognition Performance

General Considerations for Subscribers

You may want to tell subscribers about the following general considerations when they use the voice recognition features. See the User Guide for the Cisco Unity Phone Interface for additional voice-recognition usage considerations. The guide is available at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/products_user_guide_list.html.

•When voice recognition is available on the system, you hear "Voice ready" after logging on to Cisco Unity. If voice recognition is unavailable, you do not hear the prompt and you cannot use voice commands for the duration of the call. You may be able to use voice recognition if you hang up and try again in a few moments.

•After you say a key or option name, you must pause to allow Cisco Unity to detect that you are done with the entry. Because of the pause, the system may seem slower when you say options than when you press keys for options.

•When entering multi-digit numbers, you must say each digit individually. For example, to enter extension 3200, say "Three-two-zero-zero." Cisco Unity does not recognize "Thirty-two hundred" or "Three thousand, two hundred." For long extensions or phone numbers, you can say the entire number before pausing, but you must say each digit in the number.

•You cannot say multi-digit shortcuts to jump ahead in the menus. For example, you cannot say "Four-one-two" to turn on the alternate greeting. Instead, you must say the single key number for the first menu option, then pause for the system to recognize your entry before you say the key number for the option in the next menu. (You can enter menu shortcuts by pressing keys.)

•When saying option names, you can only say one option name at a time, and you can only say the name of an option in the current menu. For example, you cannot say "alternate" to enable the alternate greeting until you have first reached the Setup menu and then the Greetings menu.

•It is possible to press keys and to say key numbers or voice commands for any menu option. For example, you could choose an option from one menu by pressing a key, then choose an option from the next menu by saying the number, then choose an option from the next menu by saying the option name. However, if you press the key and say the number or name for a single option, the key press takes precedence over the spoken information. For example, if you say "One-pound," then press "4-#" while entering a month, Cisco Unity will recognize April as the month selected rather than January.

Custom Key Map Conversations

•Subscribers who are assigned to a Custom Key Map conversation are not prevented from saying the option name to access undefined or restricted conversation menus when using voice recognition. For example, if the conversation is customized to remove Setup Options, subscribers cannot press or say "Four" for this option, but they can say "Setup Options" to access this menu.

Improving Voice Recognition Performance

•Subscribers may find that voice recognition is less accurate when used in conditions where there is background noise—for example, if a subscriber is using a speaker phone or a mobile phone. In these situations, subscribers can either use key presses to enter commands or temporarily disable voice recognition for the duration of the call. Refer subscribers to the "Using Voice Recognition" chapter of the User Guide for the Cisco Unity Phone Interface available at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/products_user_guide_list.html.

•If a subscriber voice name is frequently not recognized, consider using alternate names to add phonetic spellings of hard-to-pronounce names. For example, you could add "Goolay" as an alternate name for the last name "Goulet." For information on managing alternate names, see the "Managing the Alternate Names List" section.

Disabling Voice Recognition for All Subscribers in Cisco Unity 8.x

If you want to disable voice recognition for all subscribers, do the following procedure. (Alternatively, you can use the Bulk Edit tool available in Tools Depot to disable voice recognition for all subscribers.)

To Disable Voice Recognition for All Subscribers

Step 1 In the Cisco Unity Administrator, go to the System > Voice Recognition > ASR Servers page.

Step 2 On the Voice Recognition Servers page, check the Delete check box corresponding to the server.

Step 3 Click the Save icon.

Step 4 If the Cisco Unity system is configured for failover, repeat this procedure to disable voice recognition for all subscribers on the secondary server.

	System Administration Guide for Cisco Unity Release 8.x (With Microsoft Exchange)
	Managing Voice-Recognition Features in Cisco Unity 8.x
System Administration Guide for Cisco Unity Release 8.x (With Microsoft Exchange) Index Preface Accessing and Using the 8.x Cisco Unity Administrator Managing 8.x Cisco Unity Administrator Accounts Call Management Overview in Cisco Unity 8.x Managing Call Handlers in Cisco Unity 8.x Managing Directory Handlers in Cisco Unity 8.x Managing Inteview Handlers in Cisco Unity 8.x Managing Restriction Tables in Cisco Unity 8.x Messaging and Default Accounts Overview in Cisco Unity 8.x Managing Account Policy Settings in Cisco Unity 8.x Managing Languages in Cisco Unity 8.x Managing Systemwide Settings in Cisco Unity 8.x Managing Phone View Features and Remote Message Monitor in Cisco Unity 8.x Managing Network Settings in Cisco Unity 8.x Cisco Unity 8.x Conversation Overview Managing Systemwide Conversation Settings in Cisco Unity 8.x Managing Conversation Settings That Are Controlled by Class of Service in Cisco Unity 8.x Managing Conversation Settings That Are Controlled by Subscriber or Subscriber Template Settings in Cisco Unity 8.x Managing Voice-Recognition Features in Cisco Unity 8.x Managing Classes of Service in Cisco Unity 8.x Managing Subscriber Templates in Cisco Unity 8.x Managing Subscriber Accounts in Cisco Unity 8.x Setting Up Features That Are Controlled by Subscriber or Subscriber Template Settings in Cisco Unity 8.x Setting Up Message Notifications in Cisco Unity 8.x Setting Up Broadcast Messaging in Cisco Unity 8.x Managing Distribution Lists in Cisco Unity 8.x Setting Up the Cisco Unity Greetings Administrator in Cisco Unity 8.x Recording Greetings and Names in Cisco Unity 8.x Setting Up Subscriber Workstations in Cisco Unity 8.x Subscriber Orientation in Cisco Unity 8.x	Download this chapter Managing Voice-Recognition Features in Cisco Unity 8.x Download the complete book PDF Version of the Entire Guide (PDF - 5 MB) Feedback Table Of Contents Managing Voice-Recognition Features in Cisco Unity 8.x Access to Voice Recognition Features in Cisco Unity 8.x Confirming Voice Commands Based on Recognition Confidence Scores in Cisco Unity 8.x Voice Addressing in Cisco Unity 8.x Managing the Alternate Names List Updating the Name Grammar Files Voice-Recognition Considerations in Cisco Unity 8.x General Considerations for Subscribers Custom Key Map Conversations Improving Voice Recognition Performance Disabling Voice Recognition for All Subscribers in Cisco Unity 8.x Managing Voice-Recognition Features in Cisco Unity 8.x Voice-recognition features allow subscribers to interact with Cisco Unity by speaking rather than by using the phone keypad. The features are licensed by number of concurrent sessions, and require installation and configuration of voice-recognition software on a separate and dedicated server. Voice recognition is not available to outside callers. You configure voice recognition settings on the System > Voice Recognition pages in the Cisco Unity Administrator. For information on installing and configuring a separate voice-recognition server, see the applicable Installation Guide for Cisco Unity, at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/prod_installation_guides_list.html. See the following sections in this chapter for additional details about administering voice-recognition features: •Access to Voice Recognition Features in Cisco Unity 8.x •Confirming Voice Commands Based on Recognition Confidence Scores in Cisco Unity 8.x •Voice Addressing in Cisco Unity 8.x •Voice-Recognition Considerations in Cisco Unity 8.x •Disabling Voice Recognition for All Subscribers in Cisco Unity 8.x Access to Voice Recognition Features in Cisco Unity 8.x You enable access to the voice-recognition features on the applicable Conversation page for a subscriber template or an individual subscriber in the Cisco Unity Administrator, or by using the Bulk Edit utility. Subscribers who have access to voice-recognition features can turn voice recognition on or off for themselves by phone or by using the Cisco Unity Assistant. Subscribers who have voice recognition (also known as the Press-or-Say phone input style) turned on can do the following: •Say their voicemail passwords when logging on to the Cisco Unity telephone user interface. •Press the phone key, say the number, or say the name of a menu option. For example, in a menu where three corresponds to a delete action, subscribers can press three, say "Three," or say "Delete." •Say dates and times when taking actions that require them, such as entering an end date and time for the alternate greeting. •Say the names of recipients instead of spelling them with the phone keypad when addressing messages, adding names to private lists, and using the Message Locator feature. Tip To improve the accuracy of name recognition, you can enter alternate spellings for the names of individual subscribers. Alternate spellings are entered on the Alternate Name page for individual subscribers in the Cisco Unity Administrator. See the "Voice Addressing in Cisco Unity 8.x" section for more information. •Say "Hang up" at any menu to disconnect the call. Do the following "To Enable Access to Voice-Recognition Features" procedure to enable voice-recognition access for subscribers, and to turn on voice recognition, if desired. (To make the changes for a group of subscribers, you can use the Bulk Edit tool available in Tools Depot.) Note The voice recognition features require an additional server running voice-recognition software. For information on installing and configuring a separate voice-recognition server, see the applicable Installation Guide for Cisco Unity, at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/prod_installation_guides_list.html. To Enable Access to Voice-Recognition Features Step 1 In the Cisco Unity Administrator, go to the applicable page: •To modify the template that you will use to create subscriber accounts, go to any Subscribers > Subscriber Template page, and find the template that you want to modify. Then browse to the Conversation page. •To modify an existing subscriber account, go to any Subscribers > Subscribers page and find the applicable subscriber. Then browse to the Conversation page. Step 2 Under Phone Menu Options, check the Allow Access to Voice-Recognition Features check box. Step 3 Optionally, to turn on voice recognition, check the Use Press-or-Say Phone Input Style (Includes Voice Addressing) check box. Step 4 Click the Save icon. Confirming Voice Commands Based on Recognition Confidence Scores in Cisco Unity 8.x In certain situations, subscribers may want to be prompted to confirm an important action if the confidence score on their utterance falls below a certain level. For example, if the voice-recognition server matches an utterance to the "delete" action with a confidence level of 60%, Cisco Unity can prompt the user to confirm the deletion before proceeding. The Threshold to Trigger Speech Verification value allows you to specify the confidence level that triggers the confirmation. The setting applies to voice command actions for deleting messages; addressing a reply to all recipients; sending, replying to, or forwarding a message; or calling the sender of a message (live reply). The default confidence level is 85. To Configure the Threshold to Trigger Speech Verification Step 1 In the Cisco Unity Administrator, go to the System Settings > Voice Recognition > Settings page. Step 2 Under Voice Recognition Server, enter a value for the Threshold to Trigger Speech Verification setting. The range of accepted values is 0 (never prompt for confirmation) to 100 (always prompt for confirmation). Step 3 Click the Save icon. Step 4 If the Cisco Unity system is configured for failover, repeat this procedure to configure the threshold for the secondary server. Voice recognition settings are not replicated between the primary and secondary servers. Voice Addressing in Cisco Unity 8.x Cisco Unity recognizes spoken subscriber names, so subscribers do not have to spell them by saying letters. They can say subscriber names while addressing messages, adding subscribers to private lists, using Message Locator, or addressing public distribution lists. See the following sections: •Managing the Alternate Names List •Updating the Name Grammar Files For additional information on improving name recognition, see the "Voice-Recognition Considerations in Cisco Unity 8.x" section. Managing the Alternate Names List Alternate names are different versions of a subscriber name than what is listed in the corporate directory. Cisco Unity considers these names when a subscriber uses voice recognition to address messages, add subscribers to private lists, use Message Locator, or address public distribution lists. You create alternate names for individual subscribers, and cannot add or edit alternate names from the subscriber template. Also, it is important to note that any alternate names you create are stored locally and are not propagated to other Cisco Unity servers. Any alternate names that you add or modify are not active until there has been a grammar rebuild. See the "Updating the Name Grammar Files" section for more information on rebuilding the name grammar files. When creating alternate names, you might want to add common variants of a name that subscribers might use. For example, you might want to add Bill, Billy, and Will for the name William. Or, you could use this feature to add another version of an uncommon name, unusual nicknames, or maiden names. For example, if a subscriber asked Cisco Unity to address "Mary Jameson," which was the maiden name of Mary Brown, Cisco Unity could reference this information and add the correct subscriber to the message. You could also use alternate names to add phonetic spellings of hard-to-pronounce names. For example, you could add "Goolay" as an alternate name for the last name "Goulet." Note that name recognition accuracy degrades as the number and variability of names increase. In addition, name recognition tends to be less accurate for non-English names. For this reason, we recommend that you include only the most commonly-known alternate names for subscribers. See the following procedures: •To Add Alternate Names •To Edit Alternate Names Note that you cannot add or edit alternate names on a subscriber template. To Add Alternate Names Step 1 In the Cisco Unity Administrator, go to any Subscribers > Subscribers page and find the applicable subscriber. Then browse to the Alternate Names page. Step 2 Under Add, in the Alternate Names fields, enter the alternate names. Step 3 Click Add. Step 4 Repeat Step 2 and Step 3 until all alternate names have been added. To Edit Alternate Names Step 1 In the Cisco Unity Administrator, go to any Subscribers > Subscribers page and find the account that you want to change. Then browse to the Alternate Names page. Step 2 Do any of the following: •In the Alternate Names fields, enter changes to the already-existing alternate names. •If you want to delete an alternate name, check the check box next to the name. Step 3 Click the Save icon. Updating the Name Grammar Files The name grammar files should be updated immediately after Cisco Unity installation and configuration, and updated routinely anytime Cisco Unity subscribers are created, added or imported to the Cisco Unity directory. The names of new Cisco Unity subscribers and their alternate names will not be immediately available to the voice-recognition system unless the grammar files are updated. By default, the name grammar files are scheduled to update automatically at 1:00 a.m. daily. To make changes available immediately or at a different scheduled time, do one of the following procedures: •To Schedule When Cisco Unity Automatically Rebuilds the Name Grammar Files •To Manually Rebuild the Name Grammar Files Note Depending on the size of the directory, name grammars may take a long time to build and may affect system performance. We recommend that you schedule the automatic rebuild to occur at off-peak times. By default, the automatic rebuild starts at 1:00 a.m. Manual updates are recommended only when troubleshooting or during the initial Cisco Unity configuration. To Schedule When Cisco Unity Automatically Rebuilds the Name Grammar Files Step 1 In the Cisco Unity Administrator, go to the System > Voice Recognition > Settings page. Step 2 Under Grammar Engine, in the Automatic Rebuild Time fields, enter the new time. By default, the automatic rebuild starts at 1:00 a.m. We recommend that you schedule the automatic rebuild to occur at off-peak times. Step 3 Click the Save icon. Step 4 If the Cisco Unity system is configured for failover, repeat this procedure to configure the threshold for the secondary server. Voice recognition settings are not replicated between the primary and secondary servers. To Manually Rebuild the Name Grammar Files Depending on the size of the directory, name grammars may take a long time to build and may affect system performance. Manual updates are recommended only when troubleshooting or during the initial Cisco Unity configuration. Step 1 In the Cisco Unity Administrator, go to the System > Voice Recognition > Settings page. Step 2 Under Grammar Engine, click Start Manual Update. Step 3 If the Cisco Unity system is configured for failover, repeat this procedure to update the name grammar files on the secondary server. Voice-Recognition Considerations in Cisco Unity 8.x See the following sections: •General Considerations for Subscribers •Custom Key Map Conversations •Improving Voice Recognition Performance General Considerations for Subscribers You may want to tell subscribers about the following general considerations when they use the voice recognition features. See the User Guide for the Cisco Unity Phone Interface for additional voice-recognition usage considerations. The guide is available at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/products_user_guide_list.html. •When voice recognition is available on the system, you hear "Voice ready" after logging on to Cisco Unity. If voice recognition is unavailable, you do not hear the prompt and you cannot use voice commands for the duration of the call. You may be able to use voice recognition if you hang up and try again in a few moments. •After you say a key or option name, you must pause to allow Cisco Unity to detect that you are done with the entry. Because of the pause, the system may seem slower when you say options than when you press keys for options. •When entering multi-digit numbers, you must say each digit individually. For example, to enter extension 3200, say "Three-two-zero-zero." Cisco Unity does not recognize "Thirty-two hundred" or "Three thousand, two hundred." For long extensions or phone numbers, you can say the entire number before pausing, but you must say each digit in the number. •You cannot say multi-digit shortcuts to jump ahead in the menus. For example, you cannot say "Four-one-two" to turn on the alternate greeting. Instead, you must say the single key number for the first menu option, then pause for the system to recognize your entry before you say the key number for the option in the next menu. (You can enter menu shortcuts by pressing keys.) •When saying option names, you can only say one option name at a time, and you can only say the name of an option in the current menu. For example, you cannot say "alternate" to enable the alternate greeting until you have first reached the Setup menu and then the Greetings menu. •It is possible to press keys and to say key numbers or voice commands for any menu option. For example, you could choose an option from one menu by pressing a key, then choose an option from the next menu by saying the number, then choose an option from the next menu by saying the option name. However, if you press the key and say the number or name for a single option, the key press takes precedence over the spoken information. For example, if you say "One-pound," then press "4-#" while entering a month, Cisco Unity will recognize April as the month selected rather than January. Custom Key Map Conversations •Subscribers who are assigned to a Custom Key Map conversation are not prevented from saying the option name to access undefined or restricted conversation menus when using voice recognition. For example, if the conversation is customized to remove Setup Options, subscribers cannot press or say "Four" for this option, but they can say "Setup Options" to access this menu. Improving Voice Recognition Performance •Subscribers may find that voice recognition is less accurate when used in conditions where there is background noise—for example, if a subscriber is using a speaker phone or a mobile phone. In these situations, subscribers can either use key presses to enter commands or temporarily disable voice recognition for the duration of the call. Refer subscribers to the "Using Voice Recognition" chapter of the User Guide for the Cisco Unity Phone Interface available at http://www.cisco.com/en/US/products/sw/voicesw/ps2237/products_user_guide_list.html. •If a subscriber voice name is frequently not recognized, consider using alternate names to add phonetic spellings of hard-to-pronounce names. For example, you could add "Goolay" as an alternate name for the last name "Goulet." For information on managing alternate names, see the "Managing the Alternate Names List" section. Disabling Voice Recognition for All Subscribers in Cisco Unity 8.x If you want to disable voice recognition for all subscribers, do the following procedure. (Alternatively, you can use the Bulk Edit tool available in Tools Depot to disable voice recognition for all subscribers.) To Disable Voice Recognition for All Subscribers Step 1 In the Cisco Unity Administrator, go to the System > Voice Recognition > ASR Servers page. Step 2 On the Voice Recognition Servers page, check the Delete check box corresponding to the server. Step 3 Click the Save icon. Step 4 If the Cisco Unity system is configured for failover, repeat this procedure to disable voice recognition for all subscribers on the secondary server.
	Terms & Conditions \| Privacy Statement \| Cookie Policy \| Trademarks