LinkedIn
Copied!

Table of Contents

Mapping entities in conversation text

Version:

Only available versions of this content are shown in the dropdown

To prevent your Pega Intelligent Virtual Assistant (IVA) from asking for information that the user has already provided, you can optionally configure entity mapping definitions for the cases configured in your channel. When the IVA creates a case during a chat session, the system applies the associated entity mapping and copies any entities extracted from the user input into the case properties.

For example, for a Car Insurance case, you can map the CarModel and CarMake entities to the .CarModel and .CarMake case properties. When these entities are detected in a chat conversation, their values are automatically assigned to the case properties and the system does not ask the user about the car make and model.
For the IVA to respond to user input by creating a case and asking questions, add a conversation to a case type for your application. For more information, see Adding a conversation to a case type .

When a case property has a value stored in it that was mapped from a detected entity, the IVA will skip over this question shape. This prevents the chatbot from asking for information that the user has already provided.

  1. In the header of Dev Studio, click the name of the application, and then click Channels and interfaces.

  2. In the Current channel interfaces section, click the icon that represents your existing Alexa, Unified Messaging, or Web Chatbot channel.

  3. In the channel that you select, click the Behavior tab.

  4. At the top of the preview console, turn on the Show analysis switch.

  5. In the preview console, in the right panel input field, enter a sample question.

    For a car insurance quote, enter: I would like to request a car insurance quote for my Ford Mustang GT.
    The analysis in the preview console shows the sentiment of the user input (positive, negative, neutral), the detected topic, the confidence in the detected topic (as a percentage), and the detected entities.
  6. In the create case section, click the Open response configuration icon.

  7. In the Response configuration window, click the Entities extraction tab.

  8. Add an entity to a case property mapping by clicking Add mapping.

  9. In the Entity list, click an entity.

    To map an entity defined for a car make, select Car Make.
  10. In the Case property list, click a case property that you want to map to the entity.

    To map the Car Make entity to a case type property called Make, select Make.
  11. To add another entity to the case property mapping, repeat steps 8 through 10.

  12. Click Submit.

Ensure that the IVA learns from the entities that you mapped to the case type properties by building the text analytics model. For more information, see Applying simulation changes to a text analytics model.

  • Extracting data from a conversation

    You can create and update entities from the preview console of the Pega Intelligent Virtual Assistant (IVA) and then update the text analytics model with this information. You then map the extracted entity values into case properties, creating a richer and more conversational experience for the end user.

  • Simulating a conversation and building a chatbot

    Before moving Pega Intelligent Virtual Assistant (IVA) to a production environment, you can verify whether the chatbot works correctly by using the preview console. You use the preview console to simulate a chatbot conversation and ensure that the IVA knows how to correctly respond to user input. Working in the preview console also improves the text analysis and artificial intelligence of the IVA channel.

Have a question? Get answers now.

Visit the Collaboration Center to ask questions, engage in discussions, share ideas, and help others.