Which ML Package Can I choose from AI Center-DU

Help Documentation
, ,
1

Hi, everyone,

I have a notary document. I want to scraping blue number from document using Document Understanding.

Notary document;

A blurred document with a visible stamp in the top right corner marked with the number 40076. (Captioned by AI)
A blurred document with a visible stamp in the top right corner marked with the number 40076. (Captioned by AI)574×765 41.3 KB

ı

Which model can I choose?

The image shows the "ML Packages" section with options to create a new package by uploading a zip file, using out-of-the-box packages, or importing ML packages from other AI center tenants in a software interface. (Captioned by AI)
The image shows the "ML Packages" section with options to create a new package by uploading a zip file, using out-of-the-box packages, or importing ML packages from other AI center tenants in a software interface. (Captioned by AI)1108×318 28.6 KB

This image shows a list of UiPath Document Understanding machine learning packages for extracting data from various document types, including details like headers and line items. (Captioned by AI)
This image shows a list of UiPath Document Understanding machine learning packages for extracting data from various document types, including details like headers and line items. (Captioned by AI)1368×871 108 KB

2

there is no Notary document Out-of-the-box model available for Notary document. You can build your custom ML skill using AI center.

if you only want to extract only one blue number you can use OCR and then use regex to get desired value. if this dose not work use generative extractor.

3

@Kismet_Tosun

  1. Better to choose document understanding model which is a common model when there is no specific model that exists
  2. Also recommend you to use modern DU rather than AI center for this as it gives more responsive training option and it uses docpath as the backend model which is more better for extraction

cheers