Table of Contents

Exploring text predictions and text analyzers


Only available versions of this content are shown in the dropdown

Text predictions and text analyzers for Pega Email Bot™ and Pega Intelligent Virtual Assistant™ (IVA) process user input and help the system find the best matching response by using natural language processing (NLP) and predictive and adaptive analytics. You can configure text predictions and text analyzers to detect topics, entities, sentiment, and language in an email, chat text message, or a voice command.

Text analyzer types

Your Pega Platform application supports text predictions and different types of text analyzers that you can add and configure on the Behavior tab in your email bot or IVA channel:

Exact match
The default simple text analyzer that exactly matches user input to a response.
Pega NLP
An advanced text analyzer that determines the best approximate match by using advanced natural language processing (NLP) and artificial intelligence.
Text prediction / iNLP
The default text prediction for the channel is associated with the iNLP text analyzer, as shown in the following figure. The text prediction/iNLP text analyzer detects topics, entities, sentiments, and language in a message by using NLP, predictive and adaptive analytics, and artificial intelligence.
Text predictions are targeted to replace text analyzers. As a best practice, configure text analytics for your conversational channels by using text predictions instead of text analyzers. For more information, see Understanding text analysis.
Text prediction and text analyzers in a channel
The Text Analyzer section in a channel

Text analytics model

Text analytics models provide algorithms, NLP, adaptive analysis, and artificial intelligence. To ensure that the text analyzer correctly detects a topic, entity, sentiment, or language, you first train the email bot or the IVA using the data, and then apply the changes to the text analytics model. For more information, see Training the model for the Email channel and Training the model for the IVA channel.

Did you find this content helpful?

Have a question? Get answers now.

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