Topic NLU Intent tab

  • Release version: Xanadu
  • Updated November 12, 2024
  • 2 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of NLU Intent tab

    The NLU Intent tab in Virtual Agent Designer enables ServiceNow users to review and refine natural language understanding (NLU) models tied to conversation topics. When a topic is linked to an NLU model and intent, this tab allows editing of utterances and their associated entities to improve topic discovery and prediction accuracy. This interface is available when the Now Assist Topics skill is enabled, providing an updated user experience compared to the legacy UI.

    Show full answer Show less

    Utterances tab

    This tab allows you to manage utterances for each intent efficiently. Key functions include:

    • Add an utterance: Enter unique utterances under 200 characters that represent user expressions for the intent.
    • Edit or duplicate utterances: Modify existing utterances or create copies to expand training data.
    • Move utterances: Reassign utterances to different intents if better suited elsewhere.
    • Delete utterances: Remove utterances that are irrelevant or redundant.
    • Define new entities: Select words in utterances to create entities, either simple or system-derived, which help the model understand key data points in user input.

    Entities added here also appear under the Associated Entities tab for comprehensive management.

    Associated Entities tab

    This tab displays all entities linked to the current intent, allowing you to review which utterances contain each entity. It supports both custom and system entities like date and time, enhancing the model’s ability to extract structured information from conversations.

    Conflicts tab

    Use this tab to identify and resolve conflicts where the same utterance appears in multiple intents. Detecting such conflicts in Virtual Agent Designer or NLU Workbench helps prevent ambiguity, thereby improving the accuracy and effectiveness of topic discovery and intent recognition.

    You can edit and refine ServiceNow NLU models from within Virtual Agent Designer. When a conversation topic is bound to an associated NLU model and intent, use the NLU Intent tab to review and edit utterances and associated entities.

    Note:
    An updated Virtual Agent Designer user interface is available when you install Now Assist in Virtual Agent and turn on the Now Assist Topics skill. This content assumes that you have activated this skill and can see the list view. If this skill is not enabled, you will see the legacy UI and topics page. For more information, see Virtual Agent Designer legacy topics page.

    If your topic uses NLU discovery, the topic NLU Intent tab appears on the Virtual Agent Designer interface. Test your topic to see the prediction results for topic discovery and the corresponding confidence scores for the results. Depending on the results, you can modify the associated utterances and entities in the NLU model for your topic intent.

    Utterances tab

    Use the Utterances tab to perform the following tasks.

    Table 1. Using the Utterances tab
    I want to... Do this
    Add an utterance Use the Add field to enter an utterance for an intent. Each utterance must be unique and contain fewer than 200 characters.
    Edit an utterance Select the Edit icon Edit icon. next to the utterance.
    Duplicate an utterance Select the Copy icon Copy icon. next to the utterance.
    Move an utterance to another intent Select the Menu icon Menu icon., and then select Move this utterance to another intent.
    Delete an utterance Select the Menu icon Menu icon., and then select Delete this utterance.
    Define a new entity
    1. Select a word in an utterance to use as an entity.
    2. In the Create New Entity picker, select the type of entity to be created, such as a simple or system-derived entity.

      For information about the types of entities that you can define, see NLU entities. For information about system entities, see NLU system entities. When you add an entity, it is also displayed in the Associated Entities tab.
    Figure 1. Example utterances on the NLU Intent tab
    The Check IT Ticket Status topic displays 24 utterances. There are two associated entities. You can add more utterances, or train and test your model.
    Figure 2. Adding entities on the Utterances tab
    Select a word or phrase in an utterance, then use the Create New Entity picker to add an entity related to it.

    Associated Entities tab

    Use the Associated Entities tab to view the entities defined for the intent. Select the name to view the utterances that the entity is associated with.

    Figure 3. Entities defined for an intent named UpdatePhoneNumber
    The Associated Entities sub-tab lists the entities associated with the intent.

    You can also associate system entities, such as date and time, with an intent. For more information about supported system entities, see NLU system entities.

    Conflicts tab

    Use the Conflicts tab to view conflicts among your intents. For example, if two different intents use the same utterance, this will appear as a conflict in both Virtual Agent Designer and NLU Workbench. Identifying and resolving conflicts improves the effectiveness of NLU topic discovery.

    Figure 4. Utterance conflict between two intents
    The Conflicts tab shows problems that may interfere with topic discovery. In this example, two separate intents contain the same utterance.