My top 10 wishlist

my top 10 wishlist:

  1. Advanced Search / search by field (e.g., author, title)

  2. Mass search and replace by field (e.g., in the journal names field, change all instances of Prof to Professional)

  3. smarter recognition of some fields such as publisher and journal. You can use lists to do this, since the number of such entities is (mostly) finite.

  4. Use big data / AI to improve recognition. You have a massive database of papers, You should be using this data to improve the app’s ability to recognize fields. For example, I’m sure that, for some papers, there are hundreds of users who have that paper in their account, and who also have the correct info. When another person adds that paper to their account, you should be able to use that data to recognize the correct info. Also, if there are 10,000 papers with the author Samuel J. Knapp, but only 100 with Samuel Knapp, when someone adds a paper with the author Samuel Knapp, and no authors with Samuel (any other letter) Knapp, why not at least suggest Samuel J. Knapp?

  5. Present the fields to the user before adding the paper to the user. IME, the recognition ability often makes mistakes (especially with books and book chapters). I always have to go behind PP and make sure the fields are correct. I wish PP displayed what it was going to add, BEFORE adding, so that I could correct (instead of making me go to the PP app and opening the paper). If all is well, then it’s just one click to tell PP ‘ok, go ahead and finish’. But its multiple clicks and keystrokes when it’s wrong.

  6. The ability to access ‘behind paywall’ databases. For example, I have subscriptions to Psycinfo, Ebsco, and Sage databases. PP should have a feature where it saves my userid and password, thereby enabling it to access papers behind those paywalls. As it is now, I must login to each of these databases every time I want to add a paper. And then, of course, the password times out, and when I want a new paper, I have to login again. PP should be able to do that for me.

  7. Mass label changes. I converted from Mendeley (yuckk!!!), and, as a result, I have 10 million labels. I wish there was an easy way to delete the ones I don’t want.

  8. A ‘these are the papers cited in this document’ category. I wish that PP had the ability to ‘connect’ with a document, and create a ‘category’, (such as folders and labels), for each such document. Lets call it ‘CITED’. Each paper would then have a CITED entry, and you could then click on that entry to see all the papers that you cited for the article you are writing. As it is now, I have to do that manually, by creating a folder, then moving each paper one by one.

  9. Email address for the lead author / contact person. You’d be surprised how accessible authors are and how willing they are to send you copies of their articles. Most often, these email addresses are right in the paper itself. PP should be able to scrape the paper for the email address. And hey, while we are at it, why not go whole hog and create a window to contact the author? Should be pretty easy to do.

  10. Search university repositories and other open access databases. There are a TON of articles sitting around in various repositories. Instead of solely trying to locate a paper from GS, it should be pretty easy to create a list of these repositories and then search them. For example, many law journals have their articles online, but GS usually only looks at heinonline.

OK, so there’s my top 10. I know you only have finite resources, but these are my pet peeves with PP. If you could add these, I’d be in heaven!!!

4 Likes

Nice list, I have some thoughts on two of them.

W.r.t. 4, I actually think crowd-sourcing is the wrong approach. Mendeley does this and the result is that there are 20+ versions of each paper because they have been imported in various ways from various sources and it’s not clear which one should be used. An authoritative list, to the extent one could be created, would be a better approach.

W.r.t. 7 there is a workaround, which they helped me with after I imported from Mendeley back in the day. I would confirm with Support before trying this, especially if you have a very large library - they may be able to create a restore point in advance or something in case things go horribly awry.

  1. Empty your trash.
  2. Make folders for all the labels you want to keep and add the respective papers to these.
  3. Trash all your papers (yes really)
  4. Click the little “clean up” button at the bottom of the labels list to remove unused labels (which at this point is all of them).
  5. Restore all your papers from the trash.
  6. Recreate the labels you wanted to keep and add the papers from the respective folders.
  7. Delete the folders you made in step 2.
1 Like

BIG THANKS, Jason, I had forgotten about #7 - I will have to get with support when I get time.

#4 - my experience with Mendeley is that they did everything half way. And, I agree that there is a big gray area. However, with some smarter AI (e.g., neural net), I do think it is possible to improve the recognition ability. That’s why I used an extreme example of 10,000 to 100 (99th percentile) differential. With all this data, there should be some cases that are clear cut. Especially when the meta data is not available, as a politician said recently, “what have you got to lose?” Seriously, though, at least the user could be offered options instead of blank fields.