Prompt tips
- The list description describes the overall contents for the list, while each property is a single description of an item that repeats in the list.
- You can use location hints to describe the target list’s position in the document. For examples of location hints, see Query Group extraction tips.
- For more information about how to write descriptions, or “prompts”, see Query Group extraction tips.
- For advanced options, see Advanced prompt configuration.
Troubleshooting
- If Sensible partially extracts a multi-page list, for example skipping pages in the list, use a different LLM engine. For more information, see the LLM Engine parameter.
Examples
Example 1
The following example shows using the List method to extract data from a restaurant menu:
-
Download the following example document:
| Example document | Download link | - Create a test document type in the Sensible app, then click the document type you created to edit it. In the document type’s Reference documents tab, upload the example document you downloaded in a previous step.
- Click the document type’s Configurations tab, create a new test configuration, and click the configuration you created to edit it.
- Click Sensible Instruct and create fields to extract data using the following table:
Field name | Method | Overall list description | Property ids and descriptions |
---|---|---|---|
dinners | List | ”dinner special menu items” | dinner description - “entree description”price - “dinner price” |
desserts | List | ”dessert special menu items” | dessert description - “dessert description”price - “dessert price” |
wines | List | ”red wines and white wines (not other drinks such as beers or liquors)“ | wine_name - “wine brand name”wine_type - “wine varietal name (not brand), for example, return ‘Red:cabernet savignon’ or ‘white:varietal not found‘“wine_description - “wine description”smallest_serving_price - “smallest wine serving size and its dollar price, formatted like ‘6 oz: 11‘“bottle_price - “price per bottle, in dollars” |
dinners
field:
