The Dialog Node lets you start a new dialog task within an existing dialog task if the user intent changes.
For example, let us consider a travel assistant that has the following three dialog tasks:
- Flight Availability,
- Book a Flight,
- Book a Hotel.
While handling a user’s intention to check for Flight Availability, your Virtual Assistant can detect a new user intention to Book a Flight. When this detection occurs, the VA can switch from checking flight availability to booking a flight, by invoking the dialog task that specifically handles the Book a Flight intent.
Key Features
The key features of dialog nodes are listed below:
- The ability to carry data from the current dialog task to the target dialog task as entity pre-assignments, described in a later section.
- When a target dialog task is invoked, the current dialog flow begins at the root intent of the new dialog task. If you write transitions for the dialog node, then the flow returns to the source dialog task after completing the target. If you do not write transitions, the dialog task ends after executing the target dialog task. This differs from an optional User Intent node, where the dialog flow processes a related user intent and then continues in the same dialog task thereafter.
Add the Node
Setting up an dialog node in a dialog task involves the following steps:
- Open the dialog task that you want to add the node to..
- Add a Dialog node in the designated place. Read more about adding nodes..
Configure the Node
Component Properties
Once you place your node on the canvas, the dialog window is displayed with the Component Properties tab selected by default.
The configurations you set up or edit in the Component Properties tab reflect in all the dialog tasks that use this node.
- On the Component Properties tab, under the General Settings, you can modify
- The Name,
- The Display Name,
- The Description.
NLP Properties
- From the NLP Properties tab, you can perform actions from the below-listed sections:
- Machine Learning settings to provide user utterances to improve the detection for this node.
- Patterns & Rules to capture this node if the user’s utterance contains these specific patterns.
- Bot Synonyms or related phrases for this node.
- Manage Context to control the flow of the conversation:
- Set Intent Preconditions: The intent will be detected only if any of the context tags specified as pre-conditions are available in the context at the time of intent detection.
- Configure Context Output: Define the tags to be set in the context when this intent is initiated. Intent name is automatically set as output context by the system.
Connections Properties
Click the Connections tab and set the transition properties to determine the node in the dialog task to execute next. You can write the conditional statements based on the values of any Entity or Context Objects in the dialog task, or you can use intents for transitions. If you have defined the Transition Property as End of Dialog, then these settings are ignored.
Note: The conditions configured here are applicable only for this instance and will not affect the node when being used in any other dialog. |
To set up component connection conditions, please follow the steps outlined in Adding IF-Else Conditions to Node Connections.
Instance Properties
Under the Instance Properties tab, you can configure the instance specific fields for the selected intent node. These settings are applicable only if you are connecting an existing intent node and only for the selected node instance and will not affect other dialog tasks that may invoke the node you are working with. Here is how to configure this section:
In the Entity Pre-Assignments section, you can pre-assign values from this session or node data to the entities required by the Dialog Task. Any remaining required values are handled by the sub-intent. (see below for details) If you want to reset the values, you need to use the keyword “null”, spaces would be ignored by the platform.
In the Entity Post-Assignments section, there is a list of entities available in the current task with an option to add custom variables. These entities are assigned with values from the linked dialogs session data. For example, context.entities. lists the entities from the linked sub-dialog for selection. The values are assigned once the linked dialog execution is complete.
Important Notes:
|
In the Transition Options section, you can define the flow to be followed once the task completes. You can choose from two options:
- Initiate the linked Dialog Task, and once complete, return to this node.
- End the current task and initiate the linked Dialog Task..
Click Advanced Controls to set up the below-listed options:
- Interruptions Behavior
- Use the task level Interruptions Behavior Setting: The VA refers to the Interruptions Behavior settings set at the dialog task level.
- Customize for this node: By configuring this option you can customize the Interruptions Behavior settings for this node.Read the Interruption Handling and Context Switching article for more information.
- Custom Tags defines tags to build custom profiles of your bot conversations. See here for more. You can add tags for the following:
- Message – Define custom tags to be added to the current message in the conversation.
- User – Define custom tags to be added to the user’s profile information.
- Session – Define custom tags to be added to the current conversation session.
Entity Pre-Assignments
You can pre-populate data for the destination dialog task by:
- Adding values to the entity keys.
- Adding additional custom keys and values to pass data from the current dialog task
Context
object to the destination dialog task as needed.
By default, entity nodes defined in the target dialog task are displayed as key/value pairs. You can then define values for associated entities or add custom keys and values as needed using session and context variables.
Transitioning from the source dialog task to the destination dialog task occurs at the runtime when the user input matches one of any linked dialog tasks. When you map more than one dialog task, you can define which dialog tasks are displayed to the end-users using conditional transitions.
When you call another dialog task, you want to carry information from the first dialog task to the next dialog task. For example, customer information collected in a Book a Flight dialog task is passed to the Book a Hotel dialog task.
You can use the mappedIntents
variable in the Context
object which holds the reference of the intent nodes as well as the context of the source dialog task as shown in the following JSON syntax:
{ 'title': 'title of the link', 'link':' url for the link' ,'postbackpayload': 'system generated payload'}
For example, a link to GetEmail
in the Context
object can be:
|
You can access and present these variables in a prompt message to a user to display a list of mapped dialog tasks with the link, or as a list of choices. You can pass the source dialog context to the target dialog task even if the source dialog task is no longer actively using the Context object.
The value for the postbackpayload
key is generated by the Platform and is used in some channels, such as Facebook and Slack, which defines when a link or choice is made for a target dialog task.
For more information, refer to Using Session and Context Variables in Tasks and the Context Object.