I am trying to make a summarization app on pickaxe using pickaxe chatbot but I think every time I paste an article or a YouTube videos link this app is not extracting the whole text, or transcript or for whatever reason can’t read the whole blog or article. So how can I make it read the whole article ? Should i use any action? Can anyone guide me on this please.
Hi @faisal_khalil the scraping mechanism in Pickaxe is limited in terms of being able to scrape specific webpages (e.g. it’s not capable of headless browser/beautiful soup scraping).
The best way around it is to connect a Make webhook and create a workflow in Make . com that scraped specific webpages and returns that data not your Pickaxe.
If you would like help with advanced connectivity, send me a DM with your project details so we can chat further!
1 Like
Yes @Ned.Malki is correct!
You can, however, control the scraping a bit. For end-users who are putting youtube links or weblinks into the end chat, they are handled in a special way. Here is a help post that explains how text-extraction works for end-user links.
Basically, it grabs the entire content of the thing (the video transcript or webpage content). Then it crams as much of the text as will fit into the maximum input length number. So increasing this number (which can be found under the configure tab) will make sure more of the content is put into the chatbot.
Hi Mike, I will remake this today and send your way along with other videos by Monday.
I wanted to bring to your attention that I think I made a typo on one user query on the forum (screenshot below). I edited it to accurately reflect that the scraper does not do headless browser and beautiful soup-style scraping.
1 Like