Up to you whether you want to limit your answer to software that is possible/realistic

you are viewing a single comment's thread
view the rest of the comments
[–] 5 points 2 hours ago (2 children)

I’m actually making it. Official 1.0 release is a few weeks out as I work on LLC/App Store approvals, but it’s pretty much there. BookPocket, a book tracker/reading log that treats sources/recommenders as actual entities and can learn whose tastes you share and prioritize your TBR based on those. I was annoyed the book tracker I used has been leaning heavily on AI recommendations, and none of the options out there offered this kind of functionality. I hope others find it useful, but at the very least, I’ve replaced the book tracker I was using with it and it’s helped me prioritize the books I actually want to read from my endless list.

  • source
  • hideshow 2 child comments
  • [–] [S] 2 points 2 hours ago (1 child)

    Does that mean it's a service that looks at what other users have read to build recommendations? Otherwise what is a source/recommender?

  • source
  • parent
  • hideshow 1 child comment
  • [–] 3 points 2 hours ago*

    No, all sources are ones you create and private to your account by default. So if a friend of mine recommends a book, I can add the book to my list along with adding a custom source for that friend/noting the reason they recommend it so I can remember to talk to them about it if I do read it.

    If that friend ends up joining BookPocket and adding me as a friend, I can then connect them to the source I created and we can set privacy settings for what’s shared between us. But sources are inherently custom and could be a person, podcast, library, etc. I was already doing this with tags in my last book tracker, like #recommendation #recommendation-sarah would go on a book that Sarah recommended to me, so I could use those tags to surface books actual humans recommended to me. This just pushes that a step further and allows me to add notes both on the person and the recommendation about why they recommended it, and open up new vectors for prioritizing/yearly stats.

    The core idea is to be able to prioritize my reading based on recommenders/sources I trust and tend to like their recommendations the most (and those I shouldn’t prioritize) rather than AI/algorithm based recommendations.

  • source
  • parent