Possible to improve data extracting from document understanding activities?

Help Solution Accelerators
,
1

Hi,

I’m currently working on a project which involves extract data from technical drawings (pdf files) in OCR format.
Document understanding resulted in such a great activity to do this and I’m able to extcrat many of the required information even if the format of the files is different.
The only “issue” I’m observing is that data extraction is not 100% accurate and sometimes the activity is retrieving for some fields some other data available on other lines from the technical drawings ¿Is there a way to make this extraction more accurate, and improve data retrieving?

Thanks in advance.

2

@yairb17109

Welcome to the community

Currently please show what have you done?

Are you using modern or classic?

What kind of data is it? Can you show some screenshots

Cheers

3

Train / Retrain the model with more number of sample documents to increase the efficiency if you are using document understanding to increase the accuracy