Casual reading requests

Hi, I’ve been using Paperpile for over 1 year now and these are more like the requests for more fluent browsing and casual reading.

  1. Auto-keyword generator or categorizer for the papers. Categorization is hard and with the recent advancements in AI, I think this is not so far off to ask. Especially, finding an interesting one in a given time is hard when you reach a certain amount of articles.
  2. Randomly select a paper from the directory/keyword or all articles.

Thank you!


Welcome to our forum, @dogancan, and thanks for the feedback! It doesn’t seem we’ve received these particular requests before, so I have opened new topics on our tracker for the team’s consideration. Significant changes for our extension and interface are on the works and will make it easier for the team to consider/work on this kind of request once implemented.

Check out our roadmap to see what’s on the pipeline, and please feel free to share any further feedback or questions you might have.

@dogancan raises a very important feature. As one’s library grows, organisation becomes very difficult, and using keywords/tags is a very simple and quick way to sort/organise and search a large library.

As a starting point, when a paper is imported, the keywords that the authors/journal chose for the paper can be added automatically as tags. Eventually, this can evolve to prompt the user to add additional tags when the paper is imported (in the browser extension).

I’m happy to report that it will be possible to label a paper when importing it using the Chrome extension in the new Paperpile @Yassine_Souilmi. We will be calling for beta testers for the new web app soon.

1 Like

I’ve been really getting into using the tagging feature when importing into Paperpile. It’s pretty neat. But, wouldn’t automatic tagging be better :slight_smile: Going back and tagging each paper one by one feels like a chore.

Thank you for your feedback on the new Paperpile extension, @dogancan. If you had automatic tagging, how would you like it to work? One possible solution for journal articles is having the option to import keywords automatically, but depending on research field, users might end up with too many irrelevant labels in their libraries.

Hi Suzanne,

I’d like to request multiple keywords, which we can then combine to conduct a thorough search. The goal is to ensure that no paper goes unread simply because it gets lost in our extensive library.

1 Like

Building on that, a simpler approach would be to identify keywords for each paper and then use a type of clustering to categorize the papers into n distinct clusters. The user can choose how many folders that they want.