A simple way to generate word clouds from PDFs, Word Documents, and text-based files.

Open in Google Colab and save a copy to use.
Note: The Colab notebook runs best with Google Chrome.

1. Single file -> Single word cloud
2. Multiple files -> Single word cloud
3. Multiple files -> Multiple word clouds + Combined word cloud

1. Run the entire file: From the menu, select Runtime then Run all
2. Click on the Upload Files button to upload your files
3. Set the appropriate parameters by toggling the Yes/No buttons in the "Settings" cell.
4. Click Run and wait for the word clouds to be generated. Wait till you see the message ### Done ### in the cell logs, or a wordcloud with the title "Combined Word Cloud".
5. Click Download to get a zip file of your word cloud images.

• Text in images cannot be read. Here is a workaround to extract text from images.
• Scanned PDFs (You know a PDF was scanned if you can't select text with your mouse when you open it normally)
  • SOTA OCR methods are still not perfect
  • OCR text recognition takes longer to run
  • *Key takeaway: If you can get a machine-generated PDF, use that, else tag your scanned PDF files properly by renaming them to end with _scanned.pdf

• Sometimes, the required packages fail to install correctly, leading to an error in the logs that says: ERROR: module 'PIL.Image' has no attribute 'Transpose'. In this scenario, go to Runtime in the menu, and select Restart and run all. This should fix the problem, and you can go through the steps outlined in the instructions.
• Upoading files using the Firefox browser has been known to go a bit wonky. This is an erratic bug in Colab itself. For the best experience, use Google Chrome.