How to build a model in form recognizer by defining keywords for field level extraction

Question

Hi All,

1. We are planning to build a custom model in form recognizer , we may need to define some keywords which should be used as anchor for extracting the values near to it. Apart from manually labelling the samples, is it possible to provide some keywords which the form recognizer should check and fetch the data nearby to it automatically
2. Also if we have 50 page document, and we need to analyze data only in the page range 1-5 , is it possible to define while invoking form recognizer through SDK/ REST?

Please advise. Thanks for your help

Answer 1

Hi @Sriramsubramaniyan Nadarajan , Thanks for using Microsoft Q&A Platform.

You can use the key-value pair feature in Form Recognizer to extract values near specific keywords. You can define the keywords as the key and the value you want to extract as the value. If you are looking something like this please visit this document to know more about this. https://learn.microsoft.com/en-us/azure/applied-ai-services/form-recognizer/concept-invoice?view=form-recog-3.0.0#key-value-pairs

Custom models support Key-value pairs: https://learn.microsoft.com/en-us/azure/applied-ai-services/form-recognizer/concept-custom-template?view=form-recog-3.0.0#model-capabilities

Regarding your second question, Form Recognizer can analyze data from a specific page range in a multi-page document. You can specify the page range by setting the "pages" parameter in the request body when invoking the REST API or SDK. For example, if you want to analyze pages 1 to 5 of a document, you can set the "pages" parameter to "1-5".

For more information, please refer this page for REST API: https://westus.dev.cognitive.microsoft.com/docs/services/form-recognizer-api-2022-08-31/operations/AnalyzeDocument

I hope this helps.

Regards,
Vasavi

-Please kindly accept the answer and vote 'yes' if you feel helpful to support the community, thanks.