UiPath DU 50 pages

Help AI Center
1

Please help me with the below scenario.

  1. I have a PDF with 50 pages.
  2. We need to extract details from a particular page.
  3. That particular page can be anywhere in the PDF.
    While datalabelling in AI Center can we can only train that particular page right ?
  4. Also beforeprocessing for DU steps we need to split that pdf and then perform the operation ?
1 Like
2
  1. Load the PDF
  • Use Read PDF Text to load and read the entire PDF.
  • Identify the page with the required information.
  1. Split the PDF
  • Use Extract PDF Page Range to extract the identified page.
  1. Process with AI Center
  • Upload the extracted page to AI Center.
  • Label the data and train the model.
  1. Document Understanding Workflow
  • Use Load DU Model to load the trained model.
  • Use Classify Document Scope to classify the document.
  • Use Data Extraction Scope to extract the information.

Regards,
Mohamed Farhan

3

@Ritaman_Baral,

Save your cost by splitting the PDF and send only the required page to DU for further processing/data labeling.

Yes, you should be as it will save your time and money.

Thanks,
Ashok :slight_smile:

4

This step is only applicable if there is an unique identifier in that particular page right ?

5

@Ritaman_Baral
yes, there should be some identifier or a fixed page number which should be passed to DU for processing.

You can identify the page by it’s text using Get PDF text activity and check page by page for specific text you are expecting. If text mates the expectations, just split that page.

Thanks,
Ashok :slight_smile:

6

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.