Configuring Web Chatbot text analyzer definitions
The Web Chatbot is a standalone widget that can be deployed on your website. This widget interacts with customers using a text interface. The Web Chatbot can be used to help customers find information, send automated responses to their queries, and to create a case for escalations.
The Web Chatbot configuration includes a list of text analyzer definitions. Text analyzer definitions are used for advanced text analysis to process user input and determine the best matching response. The definitions are based on natural language processing (NLP).
In the advanced configuration, you can specify several text analyzer definitions to provide a better response to users. User input is analyzed by using the first text analyzer definition in the list. If no match is found, the second text analyzer definition in the list is used, and so on, until a response match is found.
For each text analyzer definition, you can also specify whether to run it within the context of a case, outside of the context of a case, or both. For example, if you have a case that allows users to buy airline tickets, you can specify to run a text analyzer definition within the context of a case so that when the Wed Chatbot asks the user about when to book the flight, the response is automatically converted to a specific date.
You can add, delete, or modify each text analyzer definition in the list.
- In Designer Studio, click the name of your application and then click Channels and Interfaces.
- Click a Web Chatbot channel interface.
- Click the Behavior tab.
- In the Text Analyzer section, click Edit categories.
- Click Add category.
- Fill out the following fields:
- Category – Required. The natural language processing category for the chatbot to use as the response based on its must match, approximate match, and never match criteria.
- Must match – Optional. The comma-separated words or phrases to use as an exact match for the Web Chatbot content for the category. For example, for an initial greeting response, you can specify Hi.
- Approximate match – Optional. The comma-separated words or phrases to use as an approximate match for the Web Chatbot content for the category. For example, for an initial greeting response, you can specify: Hello or Good morning.
- Never match - Optional. The comma-separated words or phrases to never match the Web Chatbot content for the category. For example, for an initial greeting response, you can specify Bye.
- To define additional categories, repeat steps a and b.
- Click Submit.
- Optional: Select the Use advanced configuration check box. A list of defined text analyzers is displayed.
- To add a new text analyzer definition to the list:
- Click Add text analyzer.
- In the Text Analyzer type list, click Exact match or Pega NLP.
- Optional: If you selected a Pega NLP text analyzer type, in the Text Analyzer rule field, select a text analyzer rule that contains an NLP definition to use.
- Optional: To specify that the text analyzer definition is to run during a process that is specified by a case, select the Apply in context of a case check box.
- Optional: To specify that the text analyzer definition is to run outside of a process that is specified by case, select the Apply outside of case context check box.
- Click Submit.
- To modify a text analyzer in the list, click the Configuration icon.
- Optional: Enable the interaction timeout and define the number of minutes of inactivity after which the Web Chatbot times out of the user section.
- Click Save.
Published November 27, 2017 — Updated December 6, 2017