You can use Prediction Studio to define a taxonomy for topic detection in your application. The new method of defining taxonomy provides better user experience with the following enhancements (the numbers in brackets refer to the Prediction Studio example below):

• Searching for topics across complex taxonomies (1).
• Simplified topic management, for example, adding child topics, copying and pasting topics, changing a position in the taxonomy, and so on (2).
• Implementing multiple levels of topics that are embedded within the parent topic through the simplified and easy-to-navigate topic hierarchy view (3).
• Dedicated sections for various types of keywords, such as should words, and words, must words, and not words, in the interface for the taxonomy definition (4).
• Separating keywords from one another (5).
• Less erroneous keyword specification by using the Tab or Enter keys to add entries.
• Eliminating the prerequisite to create a CSV file with topics and keywords. If you already defined a taxonomy in a CSV file, you can still import it to Prediction Studio for editing (6).
• Instant taxonomy testing on real-life data (6).

Creating a taxonomy in Prediction Studio

By using taxonomies in Pega Platform, you can classify customer queries into various categories, which simplifies content management and can decrease the time that a customer service representative needs to respond to a customer.

For more information, see:

• Tutorial: Configuring a topic detection model for discovering keywords

• Creating keyword-based topic detection models