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.