Free Text Model rule form
Completing the Select Analysis tab

The text analytics capability of the Pega 7 Platform provides three types of analysis:

• Sentiment analysis

  Sentiment analysis aims to determine the positive, negative and neutral polarities of subjective sentences. By analyzing the content of a text unit, it is possible to estimate the emotional state of the author when writing the text and the effect he or she wants to have on the readers. Sentiment analysis on the Pega 7 Platform is achieved by means of lexicons and sentiment models.

  Lexicons

  Lexicons are lists of words and their associated sentiment. By identifying the sentiments of the individual words or phrases it is possible to estimate the general sentiment of a text unit. Lexicons are defined for each supported language separately and stored as decision data records.

  When configuring sentiment analysis, select one of the following lexicons:

  • pySentimentLexicon

  Sentiment models

  Sentiment models contain algorithms that can act on particular elements of a text unit as well as on the whole text.

  When configuring sentiment analysis, select one of the following models:

  • pySentimentModels
• Classification analysis

  Classification analysis aims to determine categories to which a text unit should be assigned; on the Pega 7 Platform this is achieved by means of taxonomies.

  Taxonomies

  Taxonomy is a list of terms related to a particular domain which are grouped or have hierarchical relationships. The list contains categories and terms assigned to them (for example: Safety concerns, "theft, steal, break, rob, intruder"). This help to categorize the unit of text by identifying the terms that appear in the text. Some taxonomies are provided by default in the .csv format.

  When configuring classification analysis, select one of the following taxonomies:

  • pyBankingTaxonomy
  • pyCustomerSupportTaxonomy
  • pyTelecomTaxonomy
• Entity extraction analysis

  Entity extraction analysis aims to determine and classify elements in a text unit into predefined categories such as the names of people, places, organizations, or according to the topics you specify. Entity extraction on the Pega 7 Platform is achieved by adding entity extraction models and specifying key words called topics and synonyms.

  Adding entity extraction models

  The models help to identify the names of people, places, and organizations.

  • To identify people's names, in the Identifier column type Person and in the Entities column, select pyPerson
  • To identify places, in the Identifier column type Location and in the Entities column, select pyLocation
  • To identify organizations, in the Identifier column type Organization and in the Entities column, select pyOrganization

  Adding topics and synonyms

  The topics and synonyms allow you to specify particular keywords and their synonyms that you want to identify in the analyzed text. For example:

  Topic Name: Pegasystems; Topic Synonyms: Pega, PEGA, Pegasystems

To enable and configure any type of analysis, select the respective check box.

Testing the Free Text Model rule

Create an instance of the Free Text Model rule and enable all types of analysis to test the text analytics capabilities of the Pega 7 Platform.

1. In the Free Text Model rule form, click Actions > Run to open the Test Page dialog box.
2. In the Load a text sample paste a unit of text that you want to analyze and click Execute.
3. In the Language section, you can check what is the language of the sample text.
4. In the Sentiment section, you can check the overall sentiment of the sample text:
   • positive
   • neutral
   • negative

5. In the Entities section, you can check the keywords identified in the sample text. These words are automatically identified by the text analytics engine.

   Note: Language, sentiment, and entities results are automatically generated using the text analytics engine.

6. In the Categories section, you can check the categories identified in the sample text. These categories are part of the selected taxonomy.
7. In the Topics section, you can check the topics you specified in the Add topics and synonyms section. These topics are also displayed in the Entities section.

Open topic with navigation