Skip to main content

Extract entity from text

Sophie avatar
Written by Sophie
Updated over 2 weeks ago

Definition and Usage

Identifies and extracts structured entities from unstructured text, such as names, addresses, and organizations. This command helps parse and structure information from raw text by identifying specific entity types you're interested in.


Parameter Values

Input parameters

Parameter

Description

Possible Values

Required

Options / Notes

LLM model

Select the language model that best suits your scenario

DeepSeek (64K)

Yes

Different models may have varying capabilities and costs

Text content

Enter the text content for entity extraction

Any text

Yes

Can use variables from previous steps

Entity type

Select a type of entity for extraction

All, People (names), Locations, Organizations (companies, institutions), Date & times, Products or brands, Addresses, Other entities

Yes

Determines what kind of entities will be extracted

Entity name

Please enter a specific entity to extract

Any text

No

Optional filter to extract only specific named entities

Error handling

Parameter Name

Description

Throw error & stop

When an error occurs, the action will trigger an error and stop the execution of the entire app.

Retry command

If an error occurs, the action will retry the command in an attempt to resolve the issue and continue the process.

Ignore error & continue

When an error occurs, the action will be ignored, and the workflow will continue without interruption.

Variables produced

The extracted entities will be stored into a new variable that you specify. This variable will contain structured data representing the entities found in the text.


Using Variables in Conditions

You can use the {x} notation or the variable icon to insert previously created variables into parameter fields. For example, you could use text content from a previous step as input for entity extraction. The output variable can be used in subsequent steps to process or display the extracted entities.


Notes

  • The quality and accuracy of entity extraction depend on the selected LLM model and the clarity of the input text.

  • Entity extraction works best on well-formed text with clear contextual information.

  • When selecting "All" as the entity type, the system will attempt to identify all possible entities in the text.

  • For more precise results, consider specifying both the entity type and entity name.

  • Be aware that entity extraction may incur costs depending on the platform and model used.

Did this answer your question?