Adding Semantic Extractor Resources

Create semantic extractor resources in the Administration Console

Before you begin: This procedure explains how the Exalead CloudView administrator creates ontology resources in the Administration Console so that they can be edited in the Business Console.

Go to Index > Data Processing > Semantic Processors.
Drag the Semantic Extractor processor to the list of current processors, and expand its configuration panel.
Click Create new next to Resource directory.
Enter the name of your semantic extractor resource, for example, mysemanticextractor.

Note: For more details on its configuration, see Semantic Extractor in the Exalead CloudView Configuration Guide.
Click Save and Apply.

Your new semantic extractor resource shall now be available in the Business Console, under Semantic > Resources.

Configure Semantic Extractor resources

To describe the configuration of Semantic Extractor resources in the Business Console, let’s take a use case where we want to apply semantic extraction on French postal addresses and telephone numbers.

We first define entities for all the terms that could be used to design a street location, for example, avenue|ave|av, boulevard|bd|bvd, etc.
Our aim is to generate a “street” annotation whenever one of these entities is detected in the corpus and assign it a specific display form.

Definition of an entity that will generate a “street” annotation when avenue, ave or av is detected in a document
We then define an entity for telephone prefixes “tel|t”.
We also specify a rule with a pattern matching both the address and the telephone number, and generating an addr_with_tel annotation with a given display form.

Example of semantic extractor rule definition
Go to the Test tab and test your semantic extractor resource by seeing how a sample text get annotated.

In this example the text sample matches the addr_with_tel rule so the entire sample is highlighted in blue. 10, place, 75009 and tel, also match the defined entities.