Skip to main content

         This documentation site is for previous versions. Visit our new documentation site for current releases.      

NLP outcomes database tables

Updated on June 30, 2022

Obtain information about the outcomes of natural language processing (NLP) by running reports on the database tables in which Pega Platform stores the results of text analysis. You can analyze the NLP outcomes in business intelligence (BI) tools to gain insights into your interactions with customers, or use them for auditing purposes to meet contractual requirements.

Two database tables hold the NLP outcomes. The tables have a one-to-many relationship and are linked through the summary ID:

Table nameClassDescription
pr_data_nlp_summaryData-NLP-Report-SummaryThe Summary table contains one record for the overall NLP outcome of each interaction.
pr_data_nlp_detailsData-NLP-Report-DetailsThe Details table contains records for each topic, entity, or intent detected in an NLP outcome. Each record is mapped to the summary record through the summary ID.

Summary table

Records in the Summary table contain high-level information about each interaction, such as the input text for the NLP analysis, source channel, language, and overall sentiment. The pyOutcome property in a record contains the overall NLP outcome in a binary large object (BLOB).

Summary table record, accessed from the Tracer
The summary table holds the NLP outcome for a message that starts with "I would like to cancel my subscription".

Details table

In the NLP outcome, for every topic, entity, or intent detected, there is an entry in the Details table. Each record contains the following properties:

Indicates the class to which the Details table applies. The class is always Data-NLP-Report-Details.
Indicates which sentiment was detected.
Indicates which topic, entity, or intent was detected.
Holds the record ID.
Holds the summary ID that maps the record in the Details table to the corresponding record in the Summary table.
Defines whether the outcome is a topic, entity, or intent.
Indicates whether the record is valid.
Details table record with an intent detection outcome, accessed from the Tracer
The detected intent is quit.
Details table record with an entity detection outcome, accessed from the Tracer
The detected entity type is Feature. The sentiment is negative.

Viewing NLP outcomes in the Summary database table

Review the results of the text analysis of messages that come through your email or chat channels by accessing records in the Summary database table that stores the NLP outcomes.

Typically, you retrieve the NLP outcomes data by querying the Pega Platform database from a database client. However, for educational purposes, you can view individual NLP Result records in Dev Studio.

  1. In Dev Studio, in App Explorer, search for the Data-NLP-Report-Summary class, and then click the class name.
    Result: The list displays the instances of the NLP Result records in the Summary database table. If the list does not contain any items, it means that no instances exist in your system.
    Instances of the Summary database table
    In Dev Studio, the NLP Result tab shows four instances of the Summary database table.
  2. Click an instance that you want to view.
    Result: The content of the record is displayed in XML format.

Have a question? Get answers now.

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

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best. is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us