Data Scraping from PDF with multiple pages and tables into excel

Help Studio
,
1

Hi,

I am trying to extract data from a table spanning multiple pages of a pdf file. However, the problem is I need to extract the data with specific criteria (i.e >= specific amount). once the criteria is met, the amount should be linked with the company name.

Any pointers?

For example

image
image1327×30 12.1 KB

Thanks in advance !!!

2

scrape the entire table and use filter data table to filter rows based on your condition

2 Likes
4

image
image1332×83 28.5 KB

how can I filter the column name “maximum funds approved…” into filter row as the error appeared with “the value for argument ‘column name’ is not set or is invalid”

5

Hey, @19028426d !! Can you share 1 PDF sample with us to find the best solution?
Only if you don’t have sensitive data.

But from what I understand from your question, a possible solution I would use would be to read the text with a READ PDF or OCR activity and use regex to extract the information of interest. If they satisfy the condition, I would keep the data.

I hope it helps!!!

7

Hey, @19028426d ! I looked for a more refined solution that will bring you much more security in your results. So let’s go!

Step 1 - Break your pdf into single pages so you can iterate one by one.

Step 2 - Use document understanding to extract the table from each PDF. It’s very simple, watch this 20 minute video. (UiPath Document Understanding: Extract Tables Out of PDFs - YouTube)

Step 3 - Merge all extracted tables.

Step 4 - Filter the final table with your required condition.

Take a look at the consistent result of the extraction I performed as a test:

image
image1835×579 81.7 KB

Hope this helps!!!

1 Like
8

Can your share your test file?
I am still following your steps

9

Sure!!
The .xaml:
Main.xaml (37.4 KB)

The Taxonomy:
taxonomy.json (4.3 KB)

The Sample Data:
sample data.pdf (1.1 MB)