back

by padolsey·5y ago·view on hn ↗
Hi HN. I made a tool that allows you to discover what books you are least likely to have come across or enjoyed, but are nonetheless highly rated. These books will likely challenge you if you tend to read quite narrowly. Briefly, for each book entered we find the rarest but highest rated intersections and provide those to you. The query can be summarised as: which highly rated books are least read (and enjoyed) by people who've also read the books entered?
13 comments
Yes, we need more engines like yours.

Instead, Amazon, Spotify, etc. simply steer me toward the more mainstream genre-media that I already know about.

Oh you like Grandaddy? You'll like the Pixies.

Oh you like the Pixies? You'll like Radiohead.

Oh you like the Radiohead? You'll like Nirvana.

Please, I want to go the opposite direction.

I listened to over 500 totally different genres on spotify lately, it's not difficult, it just means you have to do slightly more work than blindly listening to whatever spotify shoves you? The trick is to pick a specific genre of music -- soul, rock, whatever. Use spotify's playlists to get a rough idea of the genre and then use the radio feature to suggest more. There are good genre explorers, too -- i seem to remember one posted here that introduced me to tanci/pingtan, and managed to find some example works to listen to.
Maybe you're thinking of everynoise.com -- probably one of my favorite sites on the net.
Last.fm recommendations are great for this.
Yeah, last.fm remains significantly better than Spotify for listening reccs. Better yearly stats too.
I like the idea a lot, and the interface is very easy to use. However, I entered a Neil Gaiman Sandman book in my first four, and it showed me another Neil Gaiman book, which seems odd? But then again I didn't recognize it, so maybe the statistics really do point that way. But excluding books by the same authors you've entered might make sense.
Seems like a cool idea! Did you think about referral sales (Amazon Associates) as a way to fund the project instead of Patreon?

I have no affiliation or experience with either, just genuinely curious if you chose one over the other. Affiliate sales seems like a good fit for this kind of service.

I'm trying out Amazon affiliates yeh, tho no commissions as yet .. :/ I'm also exploring commercial licensing and API access possibilities.
I fell in love with this straight away. I'd love to be able to add my whole library tho.
Thank you for this service. What database are you using when looking for intersections?
It's a large graph database that I've prepared myself over two years, as a kind of part-time hobby. It's largely an automated scrape of metadata available on wikipedia, goodreads, ISBNdb etc.

Intersections are sourced from public ratings and reviews across the web. E.g. these books are all liked by one person so are can be made to intersect in the db - https://radicalreads.com/noname-favorite-books/

Fantastic effort, thank you very much!
Loved it, but I felt a large proportion of the 'bubble breakers for software engineers' are actually referenced quite often on hacker news. Not 'Taking charge of your fertility' though.
“Taking charge of your fertility” is basically about how to gather and analyze data about your body’s functioning, in order to achieve a desired health outcome (specifically, pregnancy).
Yeh -- I do agree, though I suppose that shows that the HN demographic is broader than the stereotype of software. The fertility one's quite fun.
Which two books did you add?

I got that recommendation too.

That’s really great! Congrats.

If you can do an Amazon/Audible integration, that would be awesome (and will let you build your recommendation system?).

From experience, I like to read the most populars/iconics books outside my perspective/pov, even though, if I had to grade them, it wouldn't be 5/5.

Maybe though, the most problematic of theses "ideological bubbles" is that people don't generally want to escape it.

Thanks! - When you say 'Amazon integration' do you mean like an import of books read so far? I think that's a great idea. I've been exploring doing that with Goodreads. Re: escaping ideological bubbles, I do agree, though I think that curiosity and gentle exposure can expand peoples' bubbles in a small enough way that it doesn't evoke ideological dissonance (especially in the realm of politics. Maybe there's a way that I can make it a gentler exposure to new ideas.
Anecdote : I put in the three first books of the Stormlight Archive series and it showed me many fantasy books, comics & BD, and mangas, I already read (I'd say 60 to 70%). I don't know what it should output for things not inherently political.
Very nice! Liked seeing Chomsky, Fromm, Tolle and Postman when I clicked the "software engineering" link. Will check out some of the other recommendations.
I added a few of my favorite books (mostly 20th century lit) and depending on the set got mostly children's books or non-English books. Not terribly helpful.
I added Ernst Jünger's Storm of Steel and Oswald Spengler's The Decline of the West -- not because I like them but rather because I was curious what would come up as result --, it showed me mostly religious books as suggestion (He Chose the Nails: What God Did to Win Your Heart) and The C Programming Language by Brian W. Kernighan.
Had a similar experience, most of my recommendations are childrens books.
Ditto here. Mostly children’s books.
Neat, thanks for making this! Another tool I would like: find books that I am unlikely to read, but, if I did read them, likely to rate highly.
I like it, even though it doesn't have my two books.

How about a browser extension that tracks the articles, videos, discussion boards, HN threads, etc one reads and in real time suggests other anti-bubble ones?

Not sure if your engine would support it, but I'd use it.

hi! What service did you use to search the books? where'd you get your data from? cool project!