Configuring a case and entity extraction in the Facebook chatbot preview console

When you simulate a conversation in the preview console for a Facebook channel, the chatbot can create a case as a response to the command that you enter. The chatbot uses text analysis to make a best guess to select a response to use. You can configure a case for the command and map entities to properties of the case type. Rebuilding the model improves the chances that the chatbot responds correctly in the future.
Before you begin: 

To use Pega Intelligent Virtual Assistant for Facebook in your application, you must obtain and install the Facebook channel component from Pega Exchange.

  1. In the preview console, enter a command or a greeting message.
    Result: A response for the command is displayed by the chatbot.
  2. Display more details about responses by turning on the Show analysis switch.
  3. In the Create case section, click the Gear icon.
  4. Optional: Configure case type for the command:
    1. Click the Response tab.
    2. In the Case type list, select the name of case type.
  5. Click the Entity extraction tab.
  6. Map extracted entities to case properties by performing the following steps for all entities that you want to map:
    1. Click Add mapping.
    2. In the Entity list, select an entity name.
    3. In the Case property list, select a case property name.
  7. Click Submit.
  8. Optional: When you make all the configuration changes in the preview console, rebuild the model by clicking Build.