Unanswered

AI Builder Document Processing - some values from Table cells are truncated & random cell values are missing

(0) Share

Report

Posted on by vbkr

I am trying to read a pdf data using custom AI Model (Unstructured version of Document Processing) and then input the data into an excel table. I trained the model with 5 documents and when i processed the documents found some of the data from the table is truncated (I have a table with 8 columns in pdf) and in some cases data is missing, saved as blank instead of column value.

Example-1: if a column value in the PDF table is 184, extracted value 18 in the excel file.

I can see the same value in the preview as well. I tried using advanced tagging mode and it still shows selected tagged value as 18 instead of 184. if I remove the tag and retag it, still not applied to the entire number (184).

Please let me know if anyone has come across this issue and what is the solution.

Categories:

AI Builder

I have the same question (0)

All responses (2)

Answers (0)

plarrue Microsoft Employee on at

Like (0)

Report
Copy link

Link copied!

Hi @vbkr ,

Would it be feasible to send me a sample file in a private message so I can try to build a model and repro your issue?
Did you use Fixed-template documents or General documents ?

Thanks

Philippe

Was this reply helpful? Yes No
vbkr 3 on at

Like (0)

Report
Copy link

Link copied!

Hi @plarrue

I will not be able to send you the sample file as it contains sensitive data. I used general document template because fixed template was giving issues (if table is spanned to multiple pages, some of the cell values are missing from pages-2 )

As I cannot share the sample file, what options do I have to to resolve this issue.

Was this reply helpful? Yes No