Text mining from PDF documents and results places

[janszky_nav]

Hello everyone,

I would like to ask for some help with a text search.

I have some long PDF files in which I would like to search for words.

Is there a way to export the places/rows results of the search?

I would like to see which page/row the results are on.

For example, the text I searched for text (like volvo).

And the results will be listed on a different page where I can see which rows of the documents contains that word „volvo” (e.g.: 2. page, 5. page, 13. page etc.)

Thank you for instance!

Tamas

[victor_palacios]

There is no easy way to do that (which I know), but why would you want to do this? Are you trying to see the context in which the word Volvo was used for instance? You could do that with a context window. See the challenge 38 of Just KNIME It to do so.

Solutions to "Just KNIME It!" Challenge 38

Hi everyone, here is my solution. It only finds the first instance of the lookup word. I also share component named “Word Window” on hub. Lookup word and window length can be configured from component.

If you want general information on how to extract from a PDF here is the webinar I ran in August:

Webinar: PDF Text Extraction using KNIME, Regex, and Python- August 17, 2022

Join @victor_palacios for the webinar " PDF Text Extraction using KNIME, Regex, and Python" on Wednesday, August 17 at 5 PM - 6 PM UTC +2 (Berlin) which is 10 AM - 11 PM UTC -5 (Chicago) HERE IS THE RECORDING PDF Extraction Webinar Slides In this webinar, we will parse PDF documents using the no-code, free tool KNIME and integrate it with code-based tools - Regex and Python. PDFs bring a number of unique challenges. For instance, how do we know if the PDF is text-based or image-based? If te…

[mlauber71]

@janszky_nav I adapted an older example (How to read multiple lines from PDF File - #6 by mlauber71) using R code to extract the text from PDF files and then search for the page and line where a special word (in this case “Procter”) would appear.

use R and KNIME to extract text from PDF file - search for page where text...

[ersy]

I want to do some string manipulation not word embeddings or other text-processing. I just made it up for the challenge like a small lab for string, characters, punctuation and word manipulation etc.,
Thanks for the webinar info.

[janszky_nav]

Dear mlauber71,
Thank you your solution.

[janszky_nav]

Dear Victor,

Thank you for your answer, it was very usful, and I saw the Webinar too.
Thank you.

May I have one more question?

If I have some pdf-s like the sample and I would like to extract text message/call data such as numbers and text.
The call text is located before the text “A jogosult ügy…”.
Whic one cell splitter the best option?

Thank you.
20221027_split_part_of_text.knwf (39.6 KB)
This is a sample pdf, unfortunatley i can’t upload the pdf file. Sorry.
KNIME_PDF_Sample_m.txt (316.4 KB)

[victor_palacios]

I would target the phrase “A jogosult ugy” and then extract 20-30 characters (or more) after it. Then get the data you want from that window.

Unfortunately, the data in the txt file seems corrupted?

To do the window targeting, I would use Regex.

Here is my cheat sheet for Regex:

Various Examples of Regex in KNIME – victor_palacios

Here is what you can use to target that text and extract words after it:

Finally, challenge 38 from Just KNIME It! may help you accomplish the same task.

[janszky_nav]

No, it isn’t corrupted. It is a pdf, i simply change the file extension. (pdf->txt)