Computed Field methods transform the output of one or more Field objects. Common use cases for computed fields include:
- Clean up raw output: If the output contains extra characters, strings, etc, you can often use a Computed Field method to strip out the unwanted data. Or, you can split or join data from different fields.
- Standardize output across configs: If you extract inconsistently formatted data from different vendors or documents, for example "6 month policy period" versus "six mo. policy duration", you can map to a common format. Consistently formatted output helps your application to handle extractions with fewer checks for corner cases.
- Add metadata: If a PDF lacks information that you want to include in the extraction, you can add it.
- Pick only selected choices from radio button groups or other groups.
The following global parameters are common to all Computed Field methods.
|id (required)||string||Sensible uses the ID as the key in the structured key/value output. In the API response, this output is in the |
|method (required)||object||The method describes the Computed Field method used to transform fields. This object's ID parameter specifies the method. The possible ID values are:|
|type||string (default: ||Specifies the type of the output value. For more information about types, see Field query object.|
Updated 5 months ago
Did this page help you?