Entering answers

Available in Classic and VPC

Enter answers to the questions entered. You can directly enter the message for the chatbot to respond with in the answer field. You can also enter various types of answers by setting conditions, user variables, and follow-up actions. If there are conversations with only the questions entered but not the answers, then the conversation remains incomplete and becomes excluded from model learning, resulting in the return of a failure message.

Set answer condition

You can set various answer conditions by using AND and OR logical operators. Up to five AND conditions, ten OR conditions, and a total of up to 50 conditions can be set.

Priority in conditions set

• The answers with set conditions have priority over answers without any set conditions.
• If multiple conditions are set, then the answer will only be returned when all conditions are met.
• If two or more answers meet the conditions, then a random answer that meets the question will be returned. For example, if December 25 is a Tuesday while Answer 1 is set with the [Tuesday] condition, and Answer 2 is set with the [Dec 25] date condition, both answers meet the condition. The answer returned will be a random selection between Answers 1 and 2.
• A failure message may be returned if there are no answers that meet the conditions.

Time/Day conditions

• Time condition: You can set an answer to only be used as a response during a specific time period.
• Day condition: You can set an answer to only be used as a response on a specific day of the week. You can choose multiple days.
• Date condition: You can set an answer to only be used as a response on a specific date.

Entity condition

You can set an answer to only be used as a response when the user utters or doesn't utter a specific entity. Please refer to Entity for how to add entities.

Form conditions

You can set a condition to remember the button user selected in a specific multiple-choice form and only respond to the user who selects or doesn't select a specific button.

User variable

Conditions can be branched based on the values entered in user variables. A string or number can be entered as the base value for a conditional branch. Also, you can use entities or other user variables by entering @ or *.

Custom V1.0

You can set custom conditions for an answer by calling an API. Enter the API URL to validate the conditions. Answers will only be returned when the {"valid":true} value is returned from the API URL entered.
Items returned upon request are as follows.

HttpMethod: GET
Http Headers: {
  X-KAA-USERKEY=4d4cf7425f5b4769807fb4cba66bd987,
  X-KAA-USERID=9ff4-b49e74ea22e4
}
body ={}

The items that must be returned are as follows.

{
  "valid": true  
}

Custom V2.0

You can set custom conditions for an answer by calling an API. Enter the API URL to validate the conditions. Answers will be returned only when specific string value is returned from the API URL entered. User variable values can be received when responding to custom condition V2.0.
Items returned upon request are as follows.

{
  "userId": "string",
  "userKey": "string",
  "userVariables": {
    "people": {
      "value": "17",
      "typ": "NUMBER"
    },
    "name": {
      "value": "[default] olive (end of user variable)",
      "typ": "TEXT"
    },
    "age": {
      "value": "30",
      "typ": "NUMBER"
    },
    "date": {
      "value": "{\"year\":2020, \"month\":6, \"day\":5, \"name\":\"olive\"}",
      "typ": "JSON"
    },
    "time": {
      "value": "{\"hour\":17, \"minute\":6}",
      "typ": "JSON"
    }
  }
}

The items that must be returned are as follows.

{
  "valid": "string",
  "userVariable": [
    {
      "name": "clientName",
      "value": "Gildong Hong",
      "type": "TEXT",
      "action": "EQ",
      "valueType": "TEXT"
    }
  ]
}

Voice settings

• Conditions can be checked based on the voice settings.
• Conditions can be branched by judging voice's volume, speed, pitch, and whether the alpha value is minimum/maximum.

For example, create a conversation using the voice settings and voice setting conditions as follows to design a scenario where the voice volume is adjusted according to the user's request to the chatbot, such as "Please increase the volume" "Please speak more slowly." The example is based on the case of increasing the volume value, and the detailed conditions can be set in the same way for the speed, pitch, and alpha values.

1. Create "Volume increase conversation."
2. . Register utterances such as "Please increase the volume" or "Speak up" to the questions.
   
3. Register "Increasing the volume" to the answer.
4. Click [Voice] in the Answer area, and set the volume value to "Louder than now" and "+1" in the Voice settings window.
   

The voice volume value starts at 0 and can only be increased to 10 at maximum, so if the current voice volume value is 10, the chatbot can't increase the volume value more, even if the user says "Increase the volume more." In such cases, add a "Voice setting condition" and configure to return the answer "The volume is already at its maximum," if the "voice's volume" is at the maximum value.

• If "Volume ≠ maximum value," the volume can be increased more so an answer such as "Increasing the volume" can be set to be returned.
  

• If "Volume = maximum value", then the volume can't be increased any more. Therefore, an answer such as "The volume is at its maximum" can be set to be returned.
  

Input method by answer type

Various answer types are supported as follows: default answer, image answer, multi-link answer, multiple choice answer, Flex message answer, LINE sticker answer, and audio answer. If the chatbot is linked to a messenger, then answers are converted and provided in the answer component form supported by the messenger.

Default answer

Default answer is the text-type answer. You compose a sentence that will be the chatbot's answer. You can write a regular message, but you can also use special symbols to add action methods, slots, user variables, LINE emojis, etc.

• Enter $ in the answer input field to insert an action method. For how to compose action methods, refer to Action method.
• Enter @ in the answer input field to use the value saved in a "slot." For how to save values in slots, refer to Slot other settings.
• Enter * in the answer input field to use the value saved in a user variable. For how to save values in user variables, refer to User variable.
• If you'd like to provide a LINE emoji as an answer, then enter a LINE emoji code in the form of ``{0x000000}` in the answer input field. The linked messenger must be set as LINE. Refer to LINE original emoji for supported LINE emojis.
• Add URL: Enter a URL address to connect.

Enter image answer

Image answer is an answer using images.
How to enter an image answer is as follows.

1. Select the answer type as Image answer.
2. Select the image answer type.
   • Image + Text: A type that shows images and text together. Buttons that are linked to external links can be added
   • Image: A type in which only images are exposed. It can be configured so that clicking the image will open an external link
   • Click the [Modify] button to change the image answer type
3. Click Register image to enter information in the image card.
4. When the Register image window appears, add the image.
   • Upload files: Add the image file from the PC. (recommended image size: width 150 px, height 150 px) You can adjust the image's ratio for the messenger service to be linked.
     • LINE and NAVER WORKS support the ratios of 1:1 and 1.51:1
     • NAVER TalkTalk supports the ratio of 1.82:1
     • Facebook supports the ratio of 1.91:1
   • Enter URL: Add an image by entering the image URL (recommended image size: width 150 px, height 150 px) The image ratio can't be adjusted for the images added through URLs
   • Select from registered images: Select among the images registered on the domain. Up to 50 images per domain can be registered
5. Enter the image's title and description.
   • Title: A phrase to be displayed with the uploaded image
   • Image description: A description for the uploaded image. Form input is not permitted for image answers. Forms are only permitted for default and multiple choice answers
   • Add button: Button to be displayed with the image. URL addresses can be entered, and up to 3 buttons can be added
6. Click the [Add] button at the right side of the image card to add more image cards.
   • Up to 20 image cards can be added
   • If two or more image cards are added, then they will be displayed in a carousel format. (The number of image cards displayed may vary by each messenger)
   • Click the [Change order] button to change the display order of the image cards

Enter image answer

Implement image answer

Enter multi-link answer

Multi-link answer is a type which provides an answer with a button that takes you to another webpage.
How to enter a multi-link answer is as follows.

1. Select the answer type as Multi-link answer.
2. Enter an answer which can describe the URL button in the answer field.
   • Only available for the default answer and multiple choice answer types
   • Form input is not available for multi-link answers
3. Enter the button's name and the URL address to connect.
4. Click the [Add buttons] button to add more buttons.
   • Up to 10 buttons can be added (Up to 4 buttons can be added for LINE and Facebook)

Enter multi-link answer

Implement multi-link answer

Enter multiple choice answer

A multiple choice answer is an answer type where different answers are returned, based on the different choices made on a multiple choice form.
How to enter a multiple choice answer is as follows.

1. Select the answer type as Multiple choice answer.
2. Enter the form name.
   • You can enter a new name or load an already registered form
   • When creating a new form, its name must be unique and a previous form with the same name must not exist
3. Select a button type.
   • Multiple choice button: Buttons listed vertically
     • Up to 29 buttons can be added
     • Up to 4 can be added for LINE and Facebook
     • Up to 10 can be added for NAVER TalkTalk and NAVER WORKS
   • Quick reply: Round buttons listed horizontally. Up to 10 buttons can be added
4. Enter an answer that can describe the multiple choice form.
   • Form input is available for multiple choice answers (conversations with multiple turns can be designed)
5. Enter text to be displayed on multiple choice buttons or enter an entity.
   • With an entity specified, the user entering a specific entity will trigger the same action as clicking the button
6. Connect the task to be performed when the user clicks a multiple choice button or enters the entity.
   • Button action: Enter an answer that the system will use to respond when the user clicks the button
   • Conversation connection: You can connect an existing conversation so that the chatbot responds with an answer registered to the connected conversation.
     • To check the answer for the connected conversation, click the [Check answers] button on the right
     • If the conversation name you entered does not exist, then a new conversation is automatically created using that conversation name
     • As loops within scenarios are restricted, certain conversations may not be connected
   • Text answer: You can enter plain text, form, or action method
7. Select the service to link.
   • You can set up linkage with multiple services.
   • Among the services linked to the form, the button user selects is counted as "Success" in Integrated indicators, and the remaining services that are not chosen are counted as "Rejected."
   • If you enter an action method such as LINE Switcher API, LINE Pay, or NAVER Pay in a form, then you must select to connect the corresponding services to view the right indicators in Integrated indicator.

Compose multiple choice form answer

Implement multiple choice form answer

Enter audio answer

Audio answer is an answer type that plays a specific audio file. This type is only enabled if the domain type is set as CLOVA AiCall.
How to enter an audio answer is as follows.

1. Enter a question referring to Register conversation information and Register question.
2. Select Audio answer as the answer type.
3. Enter a name for the audio file in the audio name field.
4. Enter the audio URL address in the audio URL field.

• CLOVA only supports MP3 and M3U8 file formats, and AiCall only supports the WAV (8kHz, 16-bit sample, mono channel) file formats
• Registering unsupported file formats may cause a service failure

5. Select a follow-up action type.

• Wait: Wait for the user's input after the audio answer finishes playing
• Go to conversation: Go to a designated conversation after the audio answer finishes playing
• Forward to a number: Forward the call to a designated phone number after the audio answer finishes playing
• Finish call: End the call after the audio answer finishes playing

6. Click "Service status" to set the use status, and click the [Save answer] button.

Flex message answer input method

Flex message answer is an answer type that allows you to directly write answers in various layouts using the JSON editor. The Flex message answer type is only enabled if the chatbot is linked to LINE messenger or Custom. If the linkage is set with LINE messenger, then the message should be written in accordance with LINE's Flex message composition syntax.
How to enter a Flex message answer is as follows.

1. Enter a question referring to Register conversation information and Register question.
2. Select Flex answer as the answer type.
3. Enter a description for the JSON data.
   • The description entered can be displayed as the alternate text if needed. (push alarms, chat summaries, platforms that do not support Flex messages, etc.)
4. Write answers in various layouts directly in the JSON data input field using the JSON editor.
   • The JSON data entered will be forwarded to LINE messenger or Custom channel
   • You can edit sample templates provided by LINE or create a new template by clicking the [LINE template] button
   • Completed JSON data can be copied or pasted

Compose Flex message

Implement the Flex message answer type

LINE sticker answer input method

LINE sticker answer is an answer type that uses various LINE messenger emoticons. You can provide the emoticons from LINE messenger to create a chatbot that feels friendly for the users.
How to enter a LINE sticker answer is as follows.

1. Enter a question referring to Register conversation information and Register question.
2. Select LINE sticker answer as the answer type.
3. Enter the information.
   • Messenger settings: You can only select the LINE sticker answer type when the messenger is set to LINE.
   • Package ID: Enter the package ID of the sticker you wish to use
   • Sticker ID: Enter the sticker ID of the sticker you wish to use
   • Enter the package ID and sticker ID by referring to [Sticker list]

Example

No answer

"No answer" is an answer type where no answer is returned to the user. You can design a scenario with more complexity if you make good use of the "No answer" type. For example, you can design a scenario as follows using the no answer type if you want to successively check the day condition and time condition and return different answers depending on those conditions.

This is an example designed to check the day condition and time condition in order.

1. Divide a conversation to Part A and Part B, and set a day condition for Part A. Select the no answer type for its answer
2. Set Part A's follow-up action to be a redirection to Part B
3. Set a time condition in Part B, and register different answers depending on the conditions

Consecutive answer

Consecutive answer is an answer type where the chatbot responds with successive speech bubbles by combining various answer types. Consecutive answer can consist of default, image, multilink, multiple choice, or LINE sticker answers. Up to 5 speech bubbles can be added.
How to enter a consecutive answer is as follows.

1. Enter a question referring to Register conversation information and Register question.
2. Register one or more answers by referring to Register answer.
3. Click Add new speech bubble at the bottom of the answer area.
   • Image answers do not support carousel, and only one image can be added per speech bubble.
   • A multiple choice answer can be used only once in a consecutive answer.

Set follow-up action

Follow-up action is the next action for a chatbot to perform after answering the user's question. For example, you can set the chatbot to wait for the user's next input or to transfer to another conversation after answering the user's question. If the domain is an AiCall domain in a voice environment, then you can set the chatbot to connect the call to another number or end the call.

Wait

If you select "Wait," then the chatbot will wait for the user's next input. Select this option when you need to hand over the initiative of the conversation to the user.

Conversation transfer

Conversation transfer is a feature to continue conversation by going to another designated conversation without waiting for the user's next input. Select this option when you want the chatbot to keep the initiative of the conversation, and not pass it on to the user.

Other settings

Messenger settings

• You can set an answer to only be used as a response to a specific messenger. If answers for “All messengers” and answers enabled for a specific messenger are enabled at the same time, then the answer enabled for "All messengers" is not returned.
• Selecting an answer type that is not supported by a certain messenger may result in returning a failure message to the user.
• Only "default answer" and "audio answer" types are supported when the messenger condition is set to CLOVA. Also, if an answer set as "All messengers" is linked to CLOVA, only the "text answer" of each answer type is read aloud.

Marker settings

• By enabling the marker setting, you can use the answers or context of the conversation at the point when the ${`previousChatbot} built-in action method is called. If markers are disabled for certain answer speech bubbles, then that speech bubble will be excluded from the target.
• For a description about the ${`previousChatbot} built-in action method, refer to Built-in action method.

Service status
You can set the use status of the answer. If multiple answers are the Service On status, then a random answer will be returned. For example, if two or more answers to a certain question are the Service On status, the answer will be provided randomly, regardless of the order in which the answers were registered. If the user asks the same question again within the next 5 hours, then the answer that hasn't been used will be returned at random.

Voice settings
You can change voice settings for answers in an AiCall domain. Using the "Voice settings" feature allows you to adjust the speaking speed or volume of the AiCall as you want depending on the conversation's circumstance. For example, you can set the voice speed to "slow" when you want the user to confirm a bank account number or phone number, set it to "fast" when the chatbot is going through a large amount of text for a user agreement, or set the voice volume to be "increased" for a sentence to be emphasized.

Click "Voice" in the answer area when registering conversation will open the "Voice settings" window. You can use the "Settings mode" from the voice settings window to decide whether to apply the "Voice settings" specified by the user to this particular speech bubble as one-off, or to keep the settings until you change it. You can adjust the voice's loudness in "Volume," and its speed in "Speed." If you want to increase the speed, select "Faster than now" and then adjust the figure to set it to the speed you want. You can set the voice's pitch with "Pitch," and voice's tone with "Alpha."

• Settings mode: You can select the application range of the voice settings.
• Apply to this speech bubble only: The changed voice setting will only be applied to this particular speech bubble.
• Apply also to next speech bubbles: The changed settings value will be applied to all speech bubbles that come after setting the voice. The setting will be maintained until the user changes the setting the next time.
• Volume: You can set the voice volume.
  • Default volume: Set as the volume set in CLOVA AiCall
  • Maintain current volume: Keep the current voice settings
  • Louder than now: Increase the volume than set for the current voice The volume will increase from the base value. You can choose from among the range of +1, +2, +3, +4, +5, ..., +10, and the bigger the number is, the louder the volume will be compared to the current volume.
  • Quieter than now: Reduce the volume than set for the current voice. The volume will decrease from the base value. You can choose from among the range of -1, -2, -3, -4, -5, ..., -10, and the smaller the number is, the quieter the volume will be compared to the current volume.
  • Stat change: Adjust the volume with a particular number. You can choose from among the range of 0, 1, 2, 3, 4, 5, -1, -2, -3, -4, and -5, and the changed volume will correspond to the size of the number.
• Speed: You can set the voice's speed.
  • Default speed: Set as the speed set in the CLOVA AiCall
  • Maintain current speed: Keep the current voice speed
  • Faster than now: Adjust the voice speed to be faster than the current speed The volume will decrease from the base value. You can choose from among the range of -1, -2, -3, -4, -5, ..., -10, and the smaller the number is, the faster the voice will be compared to the current speed.
  • Slower than now: Adjust the voice speed to be slower than the current speed The volume will increase from the base value. You can choose from among the range of +1, +2, +3, +4, +5, ..., +10, and the bigger the number is, the slower the voice will be compared to the current speed.
  • Stat change: Adjust the speed with a particular number. You can choose from among the range of 0, 1, 2, 3, 4, 5, -1, -2, -3, -4, and -5, and the smaller the number is, the faster the speed will be. The bigger the number is, the slower the speed will be.

• Pitch: You can set the voice's pitch.
  • Default pitch: Set as the pitch set in the CLOVA AiCall
  • Maintain current pitch: Keep the current voice pitch
  • Higher than now: Adjust the pitch to be higher than the current pitch The volume will increase from the base value. You can choose from among the range of +1, +2, +3, +4, +5, ..., +10, and the bigger the number is, the higher the pitch will be compared to the current pitch.
  • Lower than now: Adjust the pitch to be lower than the current pitch The volume will decrease from the base value. You can choose from among the range of -1, -2, -3, -4, -5, ..., -10, and the bigger the number is, the lower the pitch will be compared to the current pitch.
  • Stat change: Adjust the pitch with a particular number. You can choose from among the range of 0, 1, 2, 3, 4, 5, -1, -2, -3, -4, and -5, and the smaller the number is, the lower the pitch will be. The bigger the number is, the higher the pitch will be.
• Alpha: You can set the voice tone.
  • Default tone: Set as the tone set in the CLOVA AiCall
  • Maintain current tone: Keep the current voice tone
  • Higher than now: Adjust the tone to be higher than the current voice setting. The volume will increase from the base value. You can choose from among the range of +1, +2, +3, +4, +5, ..., +10, and the bigger the number is, the higher the alpha value will be compared to the current setting.
  • Lower than now: Adjust the tone to be lower than the current setting. The volume will decrease from the base value. You can choose from among the range of -1, -2, -3, -4, -5, ..., -10, and the bigger the number is, the lower the alpha value will be compared to the current setting.
  • Stat change: Adjust the tone with a particular number. You can choose from among the range of 0, 1, 2, 3, 4, 5, -1, -2, -3, -4, and -5, and the smaller the number is, the smaller the alpha value will be, and the bigger the number is, the bigger the alpha value will be.

Follow-up actions only for AiCall domain

These are the follow-up action menus enabled only when the domain is CLOVA AiCall.

Wait > Voice recognition mode
This mode is used when you want to receive the user's questions and answers by voice. After answering, the chatbot will wait for the user's input for the specified wait time.
If no user input is received until the set wait time expires, then a no response message from Common message is sent.

Wait > Keypad input mode
This mode is used when you want to receive the user's questions and answers in numbers only. It's used when an input of passwords or customer numbers is required, for instance.

• Maximum number of digits: Enter the maximum number of digits that can be input in the keypad input mode. Keypad input mode closes when the user input reaches the set maximum number of digits, or # is entered.
• Keypad input wait time: From the moment the chatbot has completed answering or the moment the user starts a keypad input, the chatbot will wait for the user's input for the specified wait time. If no user input is received until the set wait time expires, then a failure message from Common message is sent.
• Keypad input speech bubble index: User can start a keypad input while the chatbot is responding. If a keypad input is detected while the chatbot is responding, then the chatbot will stop responding and return the answer following the user's keypad input. For example, if it is set so that keypad input is permitted from the second speech bubble, then keypad input will not be permitted while the chatbot is answering the first speech bubble. However, the keypad input is allowed after the chatbot starts answering the second speech bubble.

Wait > No selection
The chatbot will wait until the user enters something. Select this option when you need to hand over the initiative of the conversation to the user. In case of AiCall domain, you can select a wait mode. Please note that if you do not select a wait mode in an AiCall domain, then the chatbot's no response message won't be sent, even if the user input isn't received.

Forward to a number
If you select Forward to a number, then the chatbot will respond with an answer and forward the call to the set number.
Enter the number to forward the call, excluding "-". In addition to numbers, you can enter user variables and action method V2.0.

Finish call
If you choose Finish call, then the chatbot will respond with an answer and finish the session.