Configuring the entity extraction analysis

Configure the entity extraction analysis by specifying topics, entity extraction models, and rules. Use topics to specify particular keywords and their synonyms that you want to identify in the analyzed text. Entity extraction models and rules help to identify various types of named entities.

1. In the Records panel, click Decision > Text Analyzer.
2. Open the Text Analyzer rule that you want to edit.
3. Perform at least one of the following actions:
   • Add topics.

     Perform this step if you want the system to identify specific keywords and their synonyms.

     1. In the Topics section, click Add topic.
     2. In the Name field, provide the name of the topic.
     3. In the Synonyms field for the selected topic, click Edit Synonyms.
     4. Provide the synonyms of the selected topics. Separate multiple synonyms with the comma (,).
   • Add entity extraction models.

     Perform this step if you want the system to detect and categorize entities whose names are not limited by a certain pattern or a dictionary (for example, names of organizations, people, and so on).

     1. In the Entity extraction models section, click Add model.
     2. In the Entity model field, press the Down Arrow key and select an entity model. The Type field is populated automatically with the value corresponding to the selected model.

     You can define up to six models for the entity extraction analysis. The following entity extraction models are provided by default: pyLocation, pyOrganization, pyPerson.

   • Add entity extraction rules.

     Perform this step if you want the system to categorize named entities, based on a certain pattern or a set of dictionary terms that define that entity (for example, names of hardware, identification numbers, and so on).

     1. In the Entity extraction rules section, click Add rule.
     2. In the empty Entity rule field, press the Down Arrow key and select an entity extraction rule. The Type field is populated automatically with the value corresponding to the selected model.

     The following entity extraction rules are provided by default: pyAccountNumber, pyUSZipCode, pyAddress, pyCaseID, pyDate, pyEmail, pyRelationship, pySSN, and pySalutation. For more information about entity extraction rules, see PDN article Creating entity extraction rules for text analytics.

4. Click Save.

Open topic with navigation