[HN Gopher] Show HN: I scraped 3B Goodreads reviews to train a b...
___________________________________________________________________
Show HN: I scraped 3B Goodreads reviews to train a better
recommendation model
Hi everyone, For the past couple months I've been working on a
website with two main features: - https://book.sv - put in a list
of books and get recommendations on what to read next from a model
trained on over a billion reviews - https://book.sv/intersect -
put in a list of books and find the users on Goodreads who have
read them all (if you don't want to be included in these results,
you can opt-out here: https://book.sv/remove-my-data) Technical
info available here: https://book.sv/how-it-works Note 1: If you
only provide one or two books, the model doesn't have a lot to work
with and may include a handful of somewhat unrelated popular books
in the results. If you want recommendations based on just one book,
click the "Similar" button next to the book after adding it to the
input book list on the recommendations page. Note 2: This is
uncommon, but if you get an unexpected non-English titled book in
the results, it is probably not a mistake and it very likely has an
English edition. The "canonical" edition of a book I use for
display is whatever one is the most popular, which is usually the
English version, but this is not the case for all books, especially
those by famous French or Russian authors.
Author : costco
Score : 559 points
Date : 2025-11-05 17:50 UTC (2 days ago)
(HTM) web link (book.sv)
(TXT) w3m dump (book.sv)
| thinkcontext wrote:
| I'm impressed! It didn't take many books for it to start
| suggesting other books that I liked and it showed me several
| solid choices I'm adding to my queue.
| aj_hackman wrote:
| Thank you! Because of this, "The Making of Prince of Persia:
| Journals 1985-1993" by Jordan Mechner is on its way to my house.
| qingcharles wrote:
| You definitely will not regret that purchase. It's a very
| enjoyable read.
| mentos wrote:
| Here's a human recommendation if you like that you may like
| these that I've read:
|
| 1.Sid Meier's Memoir!: A Life in Computer Games -- Sid Meier
| 2.Source Code: My Beginnings -- Bill Gates 3.Build: An
| Unorthodox Guide to Making Things Worth Making -- Tony Fadell
| 4.Prince of Persia: The Journals -- Jordan Mechner 5.A Theory
| of Fun for Game Design -- Raph Koster 6.Ask Iwata: Words of
| Wisdom from Satoru Iwata, Nintendo's Legendary CEO -- Viz Media
| (Editor) 7.Control Freak: My Epic Adventure Making Video Games
| -- Cliff Bleszinski 8.Once Upon Atari: How I Made History by
| Killing an Industry -- Howard Scott Warshaw 9.Press Reset: Ruin
| and Recovery in the Video Game Industry -- Jason Schreier
| 10.Masters of Doom: How Two Guys Created an Empire and
| Transformed Pop Culture -- David Kushner
| aj_hackman wrote:
| Thank you for the recs, Masters of Doom really lit a fire
| inside of me as a young programmer. I'd also recommend "Soul
| of a New Machine" by Tracy Kidder.
| jamesponddotco wrote:
| The recommendations are pretty good; even though I only input six
| books, it was enough for it to recommend books I have on my wish
| list. Definitely going to play around some more. Plus, the
| website is super fast, very impressive.
|
| Any chance we could get an API going at some point? Are you
| planning to open source the work?
|
| I'm interested in the scrapping of Goodreads too. I'm building a
| book metadata aggregation API and plan on building a scrapper for
| Goodreads, but I imagine using a data center IP address will be a
| problem very fast. Were you scrapping from your home network?
| costco wrote:
| Thank you for the compliments :) I used 50-100 datacenter
| proxies. I just logged requests made by the iOS app with
| Charles and then recreated the headers to the best of my
| ability though the server did not seem to be very strict at
| all. Worth noting though that static residential proxies are
| not too expensive these days anyways.
|
| Re the API: The model does actually run fairly well on CPU so
| it probably wouldn't be too expensive to serve. I guess if
| there is demand for it I could do it. I think most social book
| sites would probably like to own their recommendation system
| though.
| goatsi wrote:
| Speaking of sustained scraping for AI services, I found a
| strange file on your site: https://book.sv/robots.txt. Would
| you be able to explain the intent behind it?
| costco wrote:
| I didn't want an agent to get stuck on an infinite loop
| invoking endpoints that cost GPU resources. Those fears are
| probably unfounded, so if people really cared I could
| remove those. /similar is blocked by default because I
| don't want 500000 "similar books for" pages to pollute the
| search results for my website but I do not mind if people
| scrape those pages.
| dbl000 wrote:
| I would love an API or the dataset if you could share it
| somehow! Just to play around with my own book lists.
| esafak wrote:
| It is interesting that you chose a contextual recommender when
| you would think book affinity is not very susceptible to context.
| Did you try other models too?
| skerit wrote:
| Please make this for tv series too!
| vessenes wrote:
| OK, I just added books until you told me I had too many. Fun
| idea! I have a couple of suggestions:
|
| * UI - once someone clicks "Add" you really should remove that
| item from the suggested list - it's very confusing to still see
| it.
|
| * Beam search / diversification -- Your system threw like 100
| books at me of which I'd read 95 and heard of 2 of the other 3,
| so it worked for me as a predictor of what I'd read, but not so
| well for discovery.
|
| I'd be interested in recommendations that pushed me into a new
| area, or gave me a surprising read. This is easier to do if you
| have a fairly complete list of what someone's read, I know. But
| off the top of my head, I'm imagining finding my eigenfriends,
| then finding books that are either controversial (very wide
| rating differences amongst my fellow readers) or possibly
| ghettoized, that is, some portion of similar readers also read
| this X or Y subject, but not all.
|
| Anyway, thanks, this is fun! Hook up a VLM and let people take
| pictures of their bookshelf next.
| kace91 wrote:
| (From the site) >If you visit the "intersect" page, you can
| input multiple books and find the set of users that have read
| all of those books. This can be useful for finding longer tail
| books that weren't popular enough to meet the threshold. For
| instance, if you like reading about the collapse of the Soviet
| Union, you could put in "Lenin's Tomb" and "Secondhand Time",
| and see what other books the resultant users have read.
|
| This is how filmaffinity works, which is the best
| recommendation system I've tried. They have a group of several
| dozen 'soulmates', which are users with the most similar set of
| films seen and ratings given; recommendations are other stuff
| they also liked, and you get direct access to their lists.
|
| >then finding books that are either controversial or possibly
| ghettoized
|
| Naively, I'd say the surprises are going to be better if you
| filter more different friends, rather than more controversial
| books among your friends. As in "find me a person that's like
| me only in some ways, tell me what they love". Long term this
| method is much better at exposing you to new ideas rather than
| just finding your cliques holy wars.
| idoubtit wrote:
| The "Intersect" page was useless for me. I added 15 books,
| but got no matching user. I entered a cycle of removing-
| searching, and at 10 books I had 2 users: one had read 41353
| books, and the other 85363, with no ratings...
|
| To be useful, the "Intersect" page should have:
|
| - find near matches when there is no exact match with every
| book,
|
| - ignore fake users (can any human read 80k books in many
| languages?),
|
| - do not ignore users' votes (my input was books I liked, I
| expected to find users that rated them highly).
|
| With the "Recommend" page I had the same problem as the GP,
| and all the recommendations were useless. To fix that, I
| think some features are needed:
|
| - do not list books by authors from my list (I don't need
| recommendations for them),
|
| - add a button for marking a suggested book as "disliked" (at
| the bare minimum, it should remove it from the suggestion,
| and ideally it should influence le suggestions as much as a
| "liked" book),
|
| - do not suggest several books by the same author,
|
| - add a button to hide a suggestion or show more suggestions
| (there were dozens of books I'd read but wouldn't rate high).
| costco wrote:
| What do you think the probability that someone else read 15
| books you also read is? It's very unlikely unless they are
| all staples of a genre, part of the same series, or just
| extremely popular in general. 3-5 books is how much I would
| use on that page. I have found interesting accounts of
| medievalists, people who work at think tanks, etc with it.
|
| Fake users I would agree should be filtered, but I don't
| think filtering out users who gave it a bad review is
| necessarily the intended behavior. If I put in 3 semi
| obscure Russian history books, I am presumably looking for
| someone who is an expert in Russian history to see what
| else they read. In that case I don't care if they didn't
| like one of the books or not. Approximate matches would
| require something like LSH or cosine similarity of average
| input book embedding against average embedding of read
| books of every user which I think wouldn't work well anyone
| for retrieving anyone with a moderately long interaction
| history.
| idoubtit wrote:
| I wanted to find users that loved the same kinds of
| classical novels. The core of my list was each famous
| work of famous classical writers like Dostoievsky,
| Tolstoi, Huxley and Borges. I added a few excellent
| authors, still famous but to a lesser degree, like Italo
| Calvino or Marguerite Yourcenar. I know there are many
| readers of the whole list I wrote, I could name a few
| among my friends and family.
|
| So I think the problem was not in the existence of
| similar readers, but in the way to reach them. Few people
| that read classical books log in Goodreads (I don't) and
| even fewer input what they've read over the past decades.
| comrade1234 wrote:
| I gave up on goodreads reviews. I've been burned too many times
| by highly rated books that weren't that good. If you're into
| (horny) ya romance fantasy then goodreads is great, but it's not
| for me. I haven't really found a substitute.
| jamesponddotco wrote:
| I'm not into the social aspect, so Goodreads was never an
| option, but Hardcover[1] seems like a pretty good alternative.
|
| [1]: https://hardcover.app
| owenversteeg wrote:
| Any broadly used ratings system is total garbage. Goodreads
| ratings, Google Maps ratings, Amazon reviews, Vivino for wine,
| et cetera. Even assuming the reviews are real and genuine, most
| people just aren't good at writing reviews, and the handful
| that are often have wildly different criteria than you. Someone
| already commented with one enthusiast site - and sure,
| enthusiast sites are often better than the mainstream option
| (see also: CellarTracker for wine) but honestly my advice is to
| get good at determining the quality of the thing yourself. For
| books there are a ton of hints about what you'll be getting.
| "NYT Bestseller", "xyz book club", certain publishers, who's
| quoted on the back, when was it published, who wrote it? All of
| those things can help you rapidly identify books. I personally
| dislike most modern books and prefer the "classics", so a lot
| of this is only useful as a negative signal, but even then
| there are positive signals, for example a reference to a much
| older book.
| HeinzStuckeIt wrote:
| GR is also great if you are into academic nonfiction, Classics,
| poetry, etc. The site does, after all, let you track and review
| any publication with an ISBN. What my peers and I use it for is
| worlds apart from the romance novel or LGBT young-adult book
| reviewing community that often puts GR in the news, and far
| away from all the drama that rages around genre fiction.
| noir_lord wrote:
| It has a tendency to recommend books in the same series as are
| input (putting aside that if I like a book in a series I've
| likely already read the series).
|
| It did suggest Murderbot Diaries (not on the input but a series I
| have read and did like) and an Adrian Tchaikovsky I hadn't read
| :).
| bananaflag wrote:
| Yeah the hardest problem for recommendation systems is to find
| non-Star Wars books which are like some specific Star Wars
| books and unlike some other Star Wars books. I would say it's
| AGI-complete ;)
| noir_lord wrote:
| Ironically that is one of the few uses where I've found an
| LLM to actually be useful.
|
| ChatGPT does a fairly good job at letting you negate/refine
| whatever it was you where looking for.
| dylan604 wrote:
| I would expect a recommend of Star Trek if it were AGI-
| complete just to troll
| costco wrote:
| It's explicitly trained to predict the next book read in a
| sequence, which is why you get that behavior. There's probably
| a better way for me to handle it rather than having 5 books
| from the same series tend towards the top though.
| noir_lord wrote:
| If you have the data to know the other books in a series
| maybe split the results so you have "books in series" in one
| column and "books not in a series mentioned" in the other but
| other than that it did a better job than Kindle
| recommendations which are often hilariously off the mark.
| walthamstow wrote:
| Works pretty well with cookbooks. Very cool work.
|
| One suggestion would be to make the search less strict on
| diacritics. Searching for popular cook J. Kenji Lopez Alt was
| only successful if I entered the correct O.
| NitpickLawyer wrote:
| Interesting. I tested it with sci-fi, and it definitely
| recommends good books, but not sure how accurate it is at
| surfacing the sub genres / themes. For example for [aurora -ksr,
| seveneves, project hail mary, ender's game] it gave me dune.
| Which is a great book, but not in the "first-ish contact" style I
| hoped it would be.
|
| Another thing I noticed is that it tends to recommend 2nd and 3rd
| books in a series, which is a bit so-so. If I add the first book
| in a series, I probably already read the whole series...
| 28304283409234 wrote:
| Came here to say this (recommending book 2 and 3 in a trilogy).
| Great app otherwise!
| qingcharles wrote:
| I put in a bunch of books and hit recommendations and... I'd
| already read 95% of them, so at least we know it works well!
| (checking out the other 5% now)
|
| p.s. one idea: when you click [Add] on the recommended books
| list, it should remove it from that list
|
| p.p.s. if there is a way to filter out the spam "Summary of ____"
| books, that would be good too
| jacquesm wrote:
| I have a hard time remembering titles of books I've read if
| they are not directly related to the subject matter. No problem
| remembering the content though. With movies I remember both.
| yoz-y wrote:
| It works pretty well in the sense that after inputting only a few
| quite diverse books it gave me recommendations for a lot of books
| that I've already also read and enjoyed.
|
| I would also really like a possibility to add negative signal. It
| did also recommend books that seemed interesting to me but I
| ultimately didn't like.
|
| Overall quite impressive.
| momocowcow wrote:
| Whatever I put in, it wants me to read Sapiens :_(
| oever wrote:
| Can confirm. Stallman, Torvalds, Orwell, Harari
|
| https://book.sv/#2300585,644416
| skayvr wrote:
| I've worked in recommender systems for a while, and it's great to
| see them publicized.
|
| SASRec was released in 2018 just after transformer paper, and
| uses the same attention mechanism but different losses than LLMs.
| Any plans to upgrade to other item/user prediction models?
| costco wrote:
| I'm not an expert by any means but as far as sequential
| recommendations go, aren't SASRec and its derivatives pretty
| much the name of the game? I probably should have looked into
| HSTUs more. Also this / sparse transformers in general:
| https://arxiv.org/pdf/2212.04120
| bigskydog wrote:
| Recommend OneRec which is an improvement of HSTU and it
| recently became open source
| skayvr wrote:
| There's a few alternatives, but SASRec is a good baseline for
| next-item recommendation. I'd look at BERT4Rec too. HSTU is
| definitely a strong step forward, but stays in the domain of
| ID models. HSTU also seems to rely heavily on some extra item
| information that SASRec does not (timestamps).
|
| Other models include Google's TIGER model which uses a VAE to
| encode more information about items. Similar to how modern
| text-to-voice operates.
| costco wrote:
| Thank you for the recommendations. I didn't try BERT4Rec
| because I assumed it would perform the same or worse as
| what I already had after having read
| https://dl.acm.org/doi/pdf/10.1145/3699521. The TIGER paper
| seems interesting - I definitely want to explore semantic
| IDs in general and also because I think it could allow
| including more long-tail items.
| varenc wrote:
| I love this site, and the approach! Great seeing someone making
| good use of Goodreads data.
|
| Sadly my experience with the book recommender isn't too great
| because of the 64 book limit. If I import either the most recent
| or least recent 64 book, 95% of the books it recommends to me are
| books I've read. Though it was helpful for spotting a few books
| I've read that I didn't log on Goodreads. Guess I'm pretty
| consistent.
| costco wrote:
| I think I will expand the input books limit (sadly requires
| retraining) and or the output books limit of 30.
| nsypteras wrote:
| I'm impressed it recommended so many books i've already read and
| liked! I have a big reading backlog but once it's whittled down I
| will likely come back to this. One feature request would be to
| also show a "why this is recommended" for each recommendation so
| I can further narrow down the list for what I'm looking for
| mcbrit wrote:
| I don't know. I entered, trying to be popular but at least
| slightly? opiniated:
|
| Tigana, Hyperion, A Fire Upon the Deep, Blindsight, Moby Dick
|
| and I got a list. Sure, read all that or wasn't interested for
| reasons, I added (only Neuromancer on initial recommendations):
|
| Neuromancer, VALIS, Quantum Thief, Towing Jehovah.
|
| List did not get more interesting.
|
| Book recommendations are still kind of difficult.
| mcbrit wrote:
| If I provide that list, a (real) person doesn't ask me if I've
| read the Hobbit.
| teaearlgraycold wrote:
| I don't think past liked books are nearly enough information to
| provide a good book for you today. You need a lot more
| information about the state of someone's mind.
| mcbrit wrote:
| You're talking to a dude. (in my case.) I mentioned 8 books.
|
| I won't tell you exactly what to do, but one way to do it is
| to measure your surprise with me choosing each of those 8
| books when you provide a recommendation back to me of what I
| should read next. I think I get kind of that experience
| talking to someone about books.
|
| The algorithm didn't do that.
| teaearlgraycold wrote:
| Talking to someone about books gives you so much more
| information than a book list. Their expressions, their
| accent, their energy level, their clothes, and many other
| things help to provide supplemental information.
| submeta wrote:
| Like the idea! Wondering: Weren't the early LLMs trained on data
| in Goodreads as well? I can upload and ask ChatGPT as well, and
| it will give me similar recommendations, no?
| djoldman wrote:
| Can you share the details about the Meilisearch instance? How big
| is the box and database size?
| costco wrote:
| Everything (namely Meilisearch, Postgres and the web server in
| Go) besides the model inference is running on a Hetzner server
| with a large SSD and an "AMD Ryzen 7 3700X 8-Core Processor."
| The data.ms directory is about 40GB. Once the HN traffic dies
| down I will probably move the model back to the Hetzner server
| so I don't have to pay $0.15/hour for an A4000.
| __alexander wrote:
| Care to share the scrapped data? I would love to play around with
| it.
| demaga wrote:
| I am not sure about legal side of things here, but a Kaggle
| dataset would be really cool
| costco wrote:
| Not sure if I can. At the very least book descriptions most
| likely could not be distributed. There is an academic dataset
| with around 200M reviews though:
| https://cseweb.ucsd.edu/~jmcauley/datasets/goodreads.html
| saberience wrote:
| So you're ok with stealing the data yourself but not ok with
| providing it to others, ironic.
| guelo wrote:
| I'm surprised he got that much data. Goodreads uses several
| tricks to try to stop scrapers, for example pagination only
| works up to a few pages.
| jacquesm wrote:
| They might send him a bill for use of resources.
| cjaackie wrote:
| I'm wondering about how ethical it is to load down a
| resource in this way, open to opinions. There is a mention
| "I didn't hammer down the servers" but what does that
| really even mean? The site isn't being used as intended and
| just curious how other people feel about that.
| MattGrommes wrote:
| This is cool but I'd love the option to filter out the author of
| the book you entered. I put in Shroud by Adrian Tchaikovsky and
| almost all the books are others by him, which is fine but doesn't
| really mix up the stuff I'm reading.
| nwhnwh wrote:
| I entered "Alone Together: Why We Expect More from Technology and
| Less from Each Other" and I received books about Steve Jobs,
| Harry Potter and "The Subtle Art of Not Giving a F*ck". Like
| how???
| costco wrote:
| If you want recommendations solely based on one book, please
| try the similar page: https://book.sv/similar?id=13566692
|
| These seem to fit the description you are going for better. The
| model is trained to predict the next book in the sequence.
| Those other books you listed happen to be very popular, so in
| the absence of information about you (only having 1 book), the
| model will tend to recommend those.
| BeetleB wrote:
| > Provide 3+ books for best results.
| jauntywundrkind wrote:
| Where do nice scrapes like this end up? Are there BitTorrents out
| there for scrapes like this?
|
| Honestly this would finally be the web2.0 we all wanted & hoped
| for. It's against majesty that it's all captured owned user
| content that is legally captured by essentially public message
| boards/sites.
| jimmoores wrote:
| I unexpectedly liked this. I thought the recommendations were
| actually useful.
| parkersweb wrote:
| I sadly didn't share that experience - I fed it my goodreads
| most recent - but it largely picked up on 2 or 3 series I've
| been slowly working my way through so that most of the
| recommendation list was ALL the other books in the series (and
| the spin-off series) so I didn't really get anything useful...
| dbl000 wrote:
| Echoing what everyone else has said here - awesome site, love how
| fast it was.
|
| I did notice that when I put in a single book in a series (in my
| case Going Postal, Discworld #33) that tended to dominate the
| rest of the selection. That does make sense, but I don't want
| recommendations for a series I'm already well into.
|
| Also noticed that a few books (Spycraft by Nadine Akkerman and
| Pete Langman, Tribalism is Dumb by Andrew Heaton) that I know are
| in goodreads and reviewed didn't show up in the search. I tried
| both author's name and the title of the book. Maybe they aren't
| in the dataset.
|
| It did stumble with some books more niche books (The Complete Yes
| Minister). Trying the "Similar" button gave me more books that
| were _technically_ similar because they were novelizations of
| British comedy shows, but not what I was looking for.
|
| For more common books though it lined up very well with books
| already on my wishlist!
| costco wrote:
| Yes I would say the handling of series is probably the biggest
| problem. Once my test metrics got to a point I was happy with
| and my quality spot checks passed (can I follow the models
| recommendations from one generic history book to Steven
| Runciman, also making sure popular books don't always dominate
| the results), I was ready to release because I had been working
| on this project for so long. The solution is probably using the
| transformer model to generate 100-200 candidates and then
| having a reranker on top.
| walletdrainer wrote:
| Not just series, but I seem to mostly get a list of other
| books from the same authors.
|
| The recommendations from other authors are good, but as far
| as I can tell I've read every single one of them.
|
| Continuing to aggressively add everything it recommends
| eventually does seem to result in some interesting books I
| wasn't familiar with, but I also end up with more and more
| books that are of zero interest to me.
|
| For what it's worth, I started with: Infinite
| Jest David Foster Wallace Europe Central William T.
| Vollmann Gravity's Rainbow Thomas Pynchon White
| Noise Don DeLillo One Hundred Years of Solitude Gabriel
| Garcia Marquez
|
| It is possible that there simply aren't many books like these
| in existence, so the pool of relevant recommendations gets
| exhausted fairly quickly. I'd guess trending towards
| unrelated popular books is also just a feature of the source
| data, that largely sums up my experience with goodreads
| anyway.
|
| Very cool project though. I did end up ordering a couple of
| new books, so thank you very much.
| IanCal wrote:
| Releasing is the right choice, well done with this it's
| really cool.
|
| I've only had a short play but a solution to this problem
| might be to show authors rather than books. Or select authors
| outside of the list the user has shared and then a top n
| (1,3,5) for each of those.
|
| I feel like that's how you'd recommend to someone else - type
| of book -> unknown author -> best matching few books from
| them.
|
| After that the other side would be trying to find some
| diversity (if you think I'd like author X, personally you
| might suggest three different styles of book from them rather
| than three very similar books from them)
| Peritract wrote:
| > It did stumble with ... more niche books (The Complete Yes
| Minister).
|
| If you haven't already read it, you might like Lawrence
| Durrell's Antrobus [1].
|
| [1]
| https://www.goodreads.com/book/show/759709.Antrobus_complete
| dbl000 wrote:
| I haven't read it, but I will check it out!
| mulakosag wrote:
| Going Postal is awesome. The flood of the mails and the test to
| be the post master where you would have a slide mail into the
| hole where a vicious dog is barking.
| xkbarkar wrote:
| Have nothing to add that hasn't already been commented. Like the
| entries in the add list stay. Other than that, my recommendation
| list keeps coming up with books I have already read and loved and
| I am hitting the limit :(.
|
| So filtering would be great,
|
| I have seen a few versions of the same books listed more than
| once.
|
| Loved this. Hope you get to tune it a little.
|
| Also, thank you for not ruining the site with a single popup,
| email subscription list offer, chatbot, wheelspin from hell
| anywhere.
|
| Blessings from the popup hating part of the interwebs.
| _virtu wrote:
| Hey OP I'm building a bookclub app. Do you happen to have an api
| I could plug into? I'd love to add this to our member suggestions
| section.
| androng wrote:
| I tried to import my book list with "Import goodreads" button and
| inputting https://www.goodreads.com/user/show/68515148-andrew but
| it said "import failed, see console"
| costco wrote:
| Worked for me, could be due to server being overwhelmed
|
| Here is the URL with your books:
| https://book.sv/#52752877,46049530,18437030,52480873,3260654...
| blehn wrote:
| You should filter out authors from the input books in the output.
| If liked a book by an author, surely I'd read more of their work
| if I wanted to -- recommending them isn't helpful. Along the same
| lines, I think interesting recommendations tend to be the ones
| that (1) I like and (2) I didn't expect. The more similar the
| recommendations are to the input, the more likely I already know
| them, and the more likely to create a recommendation echo
| chamber.
| honkycat wrote:
| yep, was gonna say this. Getting recommended all of the same
| books I've already read isn't great
| Semaphor wrote:
| > You should filter out authors from the input books in the
| output.
|
| No, or at least make it configurable.
|
| I'd agree for series, but not for Authors, just because I once
| read a book by someone doesn't mean I even know they have other
| stuff, the list of Authors I read and enjoyed is _very_ long.
| Ntrails wrote:
| Agree entirely - more excluding series than authors but both
| should be options.
|
| I also i need a way to describe its recommendations as "meh".
| For example, if I put Gone Girl in, I get Girl on a Train.
| Which, _personally_ , I thought was bad. I want to exclude that
| from all future rec sets, and ideally align my preferences to
| the intersection of liked A and disliked B. vOv
| sodality2 wrote:
| This is fantastic!!! I've added many results to my want-to-read
| list, they're very on-point from very few inputs. It would be
| really cool to import from a user ID, where you can choose some
| subset of your read list to inspire new suggestions, while
| excluding all books in your want-to-read and already-read lists.
| But that's an ongoing scrape to maintain, it's a cat and mouse
| game you probably don't want to start. I wonder what the legal
| status of scraped training data is... if you don't reproduce any
| of the review data I presume you're fine?
| costco wrote:
| You can import the first or last 64 books of your read, to-
| read, or currently-reading shelves if you press the "Import
| Goodreads" button and provide your Goodreads ID.
| sodality2 wrote:
| D'oh, didn't even notice that button :P Wow, that greatly
| improved the recommendations, it even found a book I wouldn't
| say is particularly related to the others but _I_ found it
| interesting-sounding. Thanks for such a cool site!!
| stevage wrote:
| This is great. would be really nice to be able to reject
| suggestions though.
| nickthesick wrote:
| I have a web app https://bookhive.buzz which is a GoodReads
| alternative based on BlueSky's protocol. I scrape all of the book
| data from Goodreads too.
|
| I would love to be able to add a recommendation system based on
| this.
| tristor wrote:
| Two bugs to know about. First, you are using a deprecated API
| call that fails in Firefox. Second, you are using an HTTP
| endpoint that fails to upgrade to HTTPS to call the GoodReads
| API, which also fails with HTTPS-Only enabled in both Chrome and
| Firefox.
|
| The idea seems good, but since I can't import my GoodReads
| successfully, it's hard for me to try
| costco wrote:
| I use `fetch` on relative endpoints so that's odd. There
| shouldn't be any external API calls on my website other than
| whatever the Cloudflare captcha uses. I also use HTTPS-only in
| Chrome and did not experience any issues. I just tested Firefox
| with HTTPS-only on/off and Safari on my phone and I was able to
| import shelves for multiple users. Are you sure that you do not
| have any privacy settings on (can you access your shelf in
| Incognito mode)?
| mscbuck wrote:
| Awesome site and speed!
|
| My advice from someone who has built recommendation systems: Now
| comes the hard part! It seems like a lot of the feedback here is
| that it's operating pretty heavily like a content based system
| system, which is fine. But this is where you can probably start
| evaluating on other metrics like serendipity, novelty, etc. One
| of the best things I did for recommender systems in production is
| having different ones for different purposes, then aggregating
| them together into a final. Have a heavy content-based one to
| keep people in the rabbit hole. Have a heavy graph based to try
| and traverse and find new stuff. Have one that is heavily tuned
| on a specific metric for a specific purpose. Hell, throw in a
| pure TF-IDF/BM25/Splade based one.
|
| The real trick of rec systems is that people want to be
| recommnded things differently. Having multiple systems that you
| can weigh differently per user is one way to be able to achieve
| that, usually one algorithm can't quite do that effectively.
| maaaaattttt wrote:
| Speaking of TF-IDF I once added it "after" the recommendations
| to downscore items that were too popular and tended to be
| recommended too much/with too many other items (think
| Beatles/iphone) and inversely for more niche items. It might be
| too costly too do depending on how you generate the
| recommendations though.
| fennec-posix wrote:
| Very neat. Even found a couple Cold War-setting books to read and
| an entire series of 6 books on the same topic, All from searching
| up Team Yankee.
|
| Thanks for the new reading list :D
| foresterre wrote:
| It seems to work decently even with just one or two titles for
| popular titles, but less so for the niche.
|
| For example, the title "Impro: Improvisation and the Theatre" by
| Keith Johnstone, linked by another article posted to HN today
| gives back the following suggestions:
|
| - Truth in Comedy: The Manual of Improvisation by Charna Halpern
| - Steve Jobs by Walter Isaacson - 1984 by George Orwell - Harry
| Potter and the Sorcerer's Stone (Harry Potter, #1) by J.K.
| Rowling - Sapiens: A Brief History of Humankind by Yuval Noah
| Harari - The Alchemist by Paulo Coelho - The Tipping Point: How
| Little Things Can Make a Big Difference by Malcolm Gladwell -
| Dune (Dune, #1) by Frank Herbert
|
| It's a bit unfortunate that all suggestions are fairly popular
| titles, which are fairly easy to find, while the unpopular or
| niche may be just as well written but a lot harder to find.
|
| Within niche topics or books, it is also usually harder to
| provide multiple similar enough titles up front.
| costco wrote:
| It's recommended that you put at least 3 books in. If you would
| like recommendations just based on one book, click the similar
| button on the book, it should take you to this page:
| https://book.sv/similar?id=297914
| simlevesque wrote:
| The How it works it way too short :) I'd love to see some
| scripts, know the hardware you use, etc...
|
| By the way you could use Summa FTS Wasm + Duckdb Wasm to have the
| same website without any backend except file hosting. Maybe even
| just Duckdb Wasm with it's FTS would be enough. Summa FTS is very
| similar to meilisearch in essence because they're both derived
| from Tantivy.
|
| https://izihawa.github.io/summa/quick-start/
| costco wrote:
| I use a Hetzner server with Ryzen 7 3700X and an SSD.
|
| I think I could get the model to work with ONNX web but it'd be
| a 2GB download so the user experience wouldn't be too great. My
| Meilisearch index is ~40GB but I don't know how much that could
| be compressed down.
|
| Here's how the similar page for books is generated, which I
| forgot to mention on the "how it works" page:
| https://gist.github.com/chris124567/8d06d64bfe827cb7f6121f93...
| spullara wrote:
| I would love to be able to filter the resulting list by removing
| certainly all books that in the same series but I think removing
| all books by authors that I have already listed would be great to
| get new things that I haven't already read. The resulting
| recommendations maybe included 1 new book for me.
| rapatel0 wrote:
| So I tried a few disparate books independently:
|
| - Guns Germs and steel - The Alchemist - The Ramayana (a few
| others)
|
| Harry Potter and the sorcerers stone came up in all of them near
| the top. :D
| costco wrote:
| > Note 1: If you only provide one or two books, the model
| doesn't have a lot to work with and may include a handful of
| somewhat unrelated popular books in the results. If you want
| recommendations based on just one book, click the "Similar"
| button next to the book after adding it to the input book list
| on the recommendations page.
| brailsafe wrote:
| In some sense, it seems to work _well_ , but the results are sort
| of nothing special and that's not what I'd personally hope for. I
| put in three books that are unrelated and got results that
| compare to a standard book store, either from the same series or
| other meme startup tech bro recommendations that I'd often
| literally see on the same shelf. I can't say it's not good,
| because obviously that's how people browse books and that's what
| you'd get from reviews, which is perhaps why I never consult
| reviews for anything.
|
| I put in Thinking in Systems and got a bunch of engineering
| management stuff which I don't care about. Deep work of course
| gave me all the rich dad poor dad, steve jobs bio, tim
| ferriswheel crap which shouldn't surprise me at all. Girl with
| the dragon tattoo gave me the rest of the series.
|
| Thematic similarity + popularity just seems boring, I'd like
| something that surfaces unusual deep cuts that I wouldn't
| necessarily find at the book store on the same shelf, but maybe
| that I could find if I went to a great library and might be out
| of print, or that I could find on libgen.
|
| With these:
|
| - Thinking In Systems: A Primer
|
| - Paddle to the Amazon: The Ultimate 12,000-Mile Canoe Adventure
|
| - The Elements of Typographic Style
|
| I was kind of hoping to at least get "Grid Systems in Graphic
| Design" or something, but mostly got Alchemist, Zen', Into the
| Wild, almost comically mainstream cuts that of course in some
| cases I've already read or could find in a Cupertino trash can,
| not that any of them are not worth reading necessarily, but very
| typical.
|
| An option to surface rarer choices that combine signals from all
| the books on the list would be neat, like in the above case, the
| least read real adventure book that somehow touches on the
| economics of places travelled through with musings about signage
| or that just happens to use a similar prose that Robert
| Bringhurst used to make print design theory not dull.
| Recommendations that only someone with a real sweaty and weird
| venn diagram of genuine personal deep interests might conjure up,
| and that a normal person might say "why the hell would I ever
| read that" but that otherwise amazing books that are just slept
| on and might never have found a market, or maybe thematically
| dissimilar+ conceptually similar in aggregate + unpopular. I'd
| like to be able to input a seed of inspiration that I haven't
| been able to find the next deeper step in, rather than all the
| books on how to start a startup in the garage I don't have. If
| it's James Hoffman's book on brewing coffee at a high level, I
| wouldn't want another YouTubers book on brewing coffee at a high
| level, I'd want the Physics of Filter Coffee, or something in an
| adjacent sphere grid / tree branch that gives me a way to pursue
| depth AND breadth but not necessarily the same book by someone
| else, or the same book with different characters. If I've found a
| seedling or a mushroom, I'd like to explore the root system of
| that fruiting body, and then at a certain point find a new
| seedling based on what I've learned so far, or the one video with
| 50 views that's somehow the best explanation of how to handle
| back-pressure in highly concurrent systems after I've realized
| that I don't know shit about concurrency, but not so deep in the
| stack that I can't bridge the gap; make the series for me.
|
| Granted, my take here might just be an indictment of reviews in
| general, or at least those sourced from a generic site like
| goodreads/amazon which is all about popularity and armchair
| criticism.
| costco wrote:
| I would agree the results are generally OK but do not feel
| magical in most cases (I think in some specific cases they do
| though). The results can be not great if you add books across
| many disciplines. For instance if you add "The Elements of
| Typographic Style" and "The Design of Everyday Things"
| (https://book.sv/#671857,18518), you do get "Grid Systems in
| Graphic Design" but under its German name "Rastersysteme fur
| die visuelle Gestaltung."
| brailsafe wrote:
| Wow, quite impressive actually, some of those additional
| entries are quite good as well, with Ellen Lupton, Steve
| Krugman, Erik Spiekermann, and then also 1984 and Mark Manson
| for some reason. Would the latter few be because they're just
| super widely read/recommended/connected regardless of genre?
|
| Great project anyway, thanks for sharing and responding to my
| stream of consciousness, I certainly didn't mean to be
| insulting and hope my tone didn't come across that way.
| costco wrote:
| Yeah, the latter are just because they are popular. If you
| have 3+ books you tend to get less random popular books
| included
| Jayakumark wrote:
| Cool work, How much did it cost to train ? Will the source for
| training be open source ?
| 6031769 wrote:
| We won't know how much it cost until after the court case.
| dbingham wrote:
| See, now this is an excellent use of LLMs (if we're going to be
| using them at all). Low stakes if it gets shit wrong, but can
| provide some really useful and surprising answers!
|
| One request, it would be nice to not have to add Goodreads, since
| I don't use it. I've love to be able to enter a couple of book
| titles or an author and just get recommendations!
| costco wrote:
| You don't have to import your Goodreads profile. You can type
| titles and authors in the box and find books to add to the list
| that way.
| giobox wrote:
| Considering how much treasure has been poured into building
| recommendation engines for just about everything online, books
| have always been very difficult for me to find recommendations
| that work. Interested to try it!
| SomeUserName432 wrote:
| I've made multiple attempts to use "AI" to find the names of
| books I've read in the past. Never managed to find a single
| one. Quite disappointed with that. I've read so many books that
| I have no idea what were called.
| majormajor wrote:
| Neat! It's a validation of the model that 75%+ of the
| recommendations are things I've read and also enjoyed, with a few
| "read, didn't like" and some more "didn't read, don't really want
| to."
|
| But I think to break the content-bubble effects to find the
| longer tail, some way to reject or blacklist things - and have
| that be taken into effect in the model - might help.
| robertritz wrote:
| To add to this Youtube afaik uses multiple models to sprinkle
| in new content alongside your usual recommendation just for
| this.
| daemonologist wrote:
| Likewise, I put in six of my favorites and had already read
| (and enjoyed) 29 of the 30 recommendations (I'll have to check
| out Blindsight by Watts). Working great but it would be cool -
| as with pretty much every recommendation algorithm ever - to
| have more of a "discovery" capability.
| cfraenkel wrote:
| FYI, on this android tablet (android v12 / FF 144.0.2), the
| 'start typing a book title...' field doesn't do anything. On the
| Mac, it brings up a list of matches to select from.
| cyrusradfar wrote:
| I think this is cool and super fast -- kudos on whatever tech you
| needed to tackle to make it so.
|
| I don't see anyone saying safety or ethics, so I'll just put it
| out there that it has some safety and ethical considerations you
| should consider.
|
| Consider "inflammatory" books and how they could be used to harm
| a group of people. Although I recognize folks post this
| "publicly", I think the intersection feature provides more than
| Goodreads.
|
| Let's say, people who have read "Mein Kampf" & "The Anarchists
| Cookbook" or some other combination that say "Antifa" to the
| current regime.
|
| I'd recommend you have a list that you consider private, always
| and allow Users to add to that list so it's more scalable. If
| folks try to intersect with anything in that list, you can warn
| that you don't allow intersection with private books.
|
| Anyway, super fun demo!
| zeroq wrote:
| It always baffled me how we censor "Mein Kampf" but we - as a
| society - are super fine with either Alex Jones shouting about
| lasers and lizard people or Joe Rogan leaving an open mic to
| people claiming there are nuclear plants and space stations
| buried under pyramids [1].
|
| Mein Kampf is absolutely terrible piece of literature not by
| it's message but by it's quality. It's exactly something I
| would expect to find in Alex Jones cell if we would sentence
| him to a year of solitary confinement.
|
| [1] just a tiny exaggeration
| colechristensen wrote:
| I don't want "safety" or "ethics" if the requirements for them
| are banning or hiding books based on somebody's ideology
| whether or not it agrees with mine.
| cyrusradfar wrote:
| I'm not asking him to hide the book recommendations, I'm
| talking about the "intersect" user feature not doxxing people
| reading them.
| colechristensen wrote:
| I mean if that's the case a person not wanting other people
| to know they're reading something probably shouldn't put
| that on a tool designed to share what you're reading for
| other people (Goodreads).
|
| I don't think anybody needs to be protected from themselves
| sharing things intentionally.
| costco wrote:
| Thank you for the kind remarks. I would say your request is
| reasonable and it's something I thought about but it's worth
| noting that if you scroll down on a Goodreads book page you can
| see all of the users who gave it 1 star, 2 stars, etc (not just
| reviews, ratings too). And in fact Goodreads does not tell
| people this but these lists include private users. So someone
| looking to cause controversy would probably just find an
| incendiary book and look for everyone who gave it 5 stars. My
| data does not include people who marked their account as
| private or only visible to those who are signed in and also
| does not give information on stars. It is also very easy to
| remove your data if you are so inclined. I only log request
| URLs but I can tell that many thousands of people visited the
| site and less than 10 opted out despite "Remove My Data" being
| prominently displayed and referenced in the attached text to my
| post.
|
| Also, I don't use them but based on forum posts I read I think
| other services like LibraryThing or Storygraph do expose
| similar information about book readers.
| the_coffee_bean wrote:
| Amazing work. Do you plan on publishing the training code?
| zeroq wrote:
| Great work!
|
| Some five years ago I was day dreaming about recommendation
| engine for movies where you could say "hey Ciri, give me a good
| gangster flick", and it will come up with something that you
| haven't seen yet but you'd definitely love.
|
| To my amazement almost everyone, even true AI believers, thought
| it was impossible to achieve. :d
|
| But my question is - having such huge dataset, do we really need
| AI for it? SASRec/RAG is sexy, but could the same result be
| achieved with simple ranking and intersections like lastfm did in
| the past with music?
|
| Some twenty years ago I came up with an idea of "brain" data
| structure for recommendations where you have all your items
| (books, movies or articles) modeled as a graph, and whenever you
| pick something it makes a ripple effect, effectively raising
| scores in cascade of every adjecent item.
|
| Just like your brain works - when you stumble upon something new
| it immediately brings back memories of similar things from the
| past. I never had the opportunity to implement it and test in
| real life scenario, but I'd be surprised if a variant of this is
| not widely used across different recommendation systems, like
| Amazon.
| esafak wrote:
| last.fm used a primitive machine learning algorithm too, else
| what are you going to rank by?
| zeroq wrote:
| Did they? I recall similar site back from as far as from
| 2008. Might be them or something similar.
|
| Anyway. I can totally see such site running purely on
| statistics. Every song, every artists, every genre is a
| bucket. You listen to a song you put a drop in these buckets.
| Once there's enough water running we can compare you to other
| users and their buckets.
|
| It might be hard to run it on scale in real time, but c'mon,
| it's leetcode junior level assignment level of complication.
| krisoft wrote:
| > To my amazement almost everyone, even true AI believers,
| thought it was impossible to achieve.
|
| I mean. Is it possible? Hey Zeroq give me a good gangster flick
| please?
|
| How will you know what i have seen and what I haven't yet? How
| will you know what movies I like? Are there even good gangster
| flicks i would enjoy and haven't seen yet?
|
| The way the problem is phrased it sounds like your dream
| recommender has two properties: "it doesn't receive any other
| information than what is in the prompt" and "it always
| recommends a movie you haven't seen and will enjoy". Those
| together are what makes your dream recommendation engine
| impossible. If you relax those then of course it is possible.
| That's just a recommendation engine.
| gen220 wrote:
| The best way I've found for finding predictably enjoyable fiction
| is to read interviews with the authors I like, and read about the
| works and authors they admire or are influenced by. Or who they
| exchanged letters or communications with, if they're long dead
| and no interviews proper exist.
|
| Strongly recommend giving that a try yourself. And trying to
| build an algorithm around it!
|
| Here's an example: Tolstoy really admired Turgenev, who was
| friends with Theodore Storm and Gustave Flaubert, and greatly
| admired Gogol.
|
| If you like Anna Karenina you'll probably find something of value
| in Torrents of Spring, Immensee, Madame Bovary or Dead Souls.
|
| It spiders out pretty quickly!
| maxglute wrote:
| Would be nice to not recommend books by author already added for
| novel discoverability.
| lifeisstillgood wrote:
| Goodreads - "hey those user written comments belong to us, you
| need to pay us"
|
| HNUser - "OpenAI told you to go swivel until they made a billion
| and you accepted that. Samesies "
| mpern wrote:
| I read the title as "I scraped 38 book reviews". Time to get
| reading glasses...
| sajb wrote:
| You could color code results from the same author or from the
| same series as an already added book, since the user most likely
| already knows about them. Perhaps a toggle to filter these out
| altogether.
| iamcreasy wrote:
| Very fast. Thanks for building it.
|
| Besides title, I'd like to provide suggestion on what type of
| books I am looking for.
| illdave wrote:
| This is really great - refreshing to have something that's
| instantly useful, with no need to signup/login. Really fast,
| immediately helpful - this is wonderful.
| easywood wrote:
| Thank you very much, I have always wished for something like this
| to be part of Goodreads itself. The intersect function especially
| will help me find hidden gems that other likeminded people have
| found. I'm looking forward to find out what books I have missed
| all my life.
| SilverSlash wrote:
| Bless this man! I despise Goodreads but continue to use it,
| because there are no real alternatives. It feels like they
| outsourced the creation of that website to some cheap consulting
| agency in a low cost location and then left it at that. For
| example, Goodreads hasn't updated its outdated version of React
| in years.
|
| For a while now I have really wanted good book recommendations
| matching my tastes. The LLMs suck at this (likely due to the mode
| collapse that Karpathy mentioned in his excellent podcast
| appearance on Dwarkesh) and Amazon is very good but only
| recommends based on the current book you're browsing.
|
| I will try this out now! But could you increase the number of
| books fed to the recommender or maybe get the top-64 highest
| rated books instead of just the most or least recent 64?
| foota wrote:
| Many years ago I built a D3 graph based explorer for movies using
| the IMDB API. You would start by entering a movie title and it
| would pull up the similar movies from IMDB and you could click
| them and see more similar ones and they would all be connected
| based on similarity. It was very fun!
| voidUpdate wrote:
| Does this break part 4 of the Goodreads TOS?
|
| "[...] you agree not to sell, license, rent, modify, distribute,
| copy, reproduce, transmit, publicly display, publicly perform,
| publish, adapt, edit or create derivative works from any
| materials or content accessible on the Service. Use of the
| Goodreads Content or materials on the Service for any purpose not
| expressly permitted by this Agreement is strictly prohibited."
|
| Also did the reviewers give you permission to fed their content
| into an LLM?
| petralithic wrote:
| Why ask questions you already know the answers to?
| hananova wrote:
| Because some tech adjacent people still have morals?
| kosolam wrote:
| I'm not taking sides in this debate, however since feeding
| whole books into LLMs is considered legal fair use now, I guess
| these reviews don't require a permission as well. Would be
| great to hear a professional lawyer take on this.
| saaaaaam wrote:
| The hidden gotcha in the Anthropic judgement (which I think
| is what you're referencing?) is that feeding whole books into
| LLMs is considered legal fair use _if you obtain them
| legitimately_.
|
| I suspect we need to wait for the NYT (and others) case to be
| decided before we know whether scraping sites in
| contravention of their terms is also fair use for LLM
| training.
|
| My own opinion (as someone who creates written content on an
| occasional professional basis) is that if you can't monetise
| your content in some other way than blocking people from
| accessing it then your content probably isn't as valuable as
| you think.
|
| But at the same time that's tricky when it's genuine
| journalism, as in NYT's case.
|
| Obviously user generated content reviewing books online is
| rather different because the motivation of the reviewers was
| (presumably) not to generate money. And, indeed, with
| goodreads there's a strong argument that people have already
| been screwed over after their good faith review submissions
| were packaged up as an asset and flogged to Amazon. A lot of
| people were quite upset by that when it happened a decade or
| so back.
|
| So from a 'moral arguments' perspective I don't think
| scraping goodreads is as problematic as other scraping
| examples.
|
| (Sorry, none of this was aimed at you - your comment just got
| me thinking and it seemed as good a place as any to put it!)
| IncreasePosts wrote:
| Goodreads offers those reviews up publicly by serving them
| from their webservers to anyone who asks for it.
| saaaaaam wrote:
| Sorry, I don't understand the point you're making. I know
| that these are publicly available - the point I was
| making, drawing off the parent comment, is that where it
| has been deemed fair use in copyright to use books to
| train LLMs _when the content has been legitimately
| obtained_ then a similar assessment might apply for this
| sort of ingestion.
|
| If content is publicly available that does not
| necessarily mean it's free of copyright control: the
| justification for using the reviews to train an LLM would
| be based on the fact that fair use means it is not an
| infringement of copyright. But if the publisher has terms
| that forbid scraping then that may mean the fair use
| argument is undermined if it is precedent in the content
| being legitimately obtained. I'm not a lawyer but it's
| quite easy to see how "books can be used for LLM training
| under fair use but not if you pirate them" extends to
| "content on the web can be used for LLM training under
| fair use but not if you've breached the terms set out by
| the publisher".
| esskay wrote:
| Fairly meaningless in this day and age. Also IIRC scraping
| legality depends heavily on jurisdiction. Some places take a
| more permissive view of accessing publicly available
| information, even if a site's TOS forbids bots.
|
| In the US there's a major precedent [0] which held that
| scraping public-facing pages isn't a CFAA "unauthorized access"
| issue. That's a big part of why we've seen entire venture-
| backed scraping companies pop up - it's not considered hacking
| if the data is already public.
|
| [0] https://en.wikipedia.org/wiki/HiQ_Labs_v._LinkedIn
| voidUpdate wrote:
| So if you are legally allowed to "adapt, edit or create
| derivative works from any materials", what's the point of the
| TOS?
| hopelite wrote:
| That's a good question. It also would not be the first time
| that companies use trickery and manipulation or even
| deliberately illegal practices for various
| business/financial reasons. At the very least it could be
| used as a tool to underpin intimidating lawsuits and
| another step up, regardless of the legality in the relevant
| jurisdiction, it could be used to influence official
| government foreign policy to exert pressure on a
| jurisdiction that permits scraping.
| zigzag312 wrote:
| I believe TOS is binding as long as it doesn't conflict
| with the law. If something is deemed fair use under the
| law, TOS cannot override those legal rights.
| hrimfaxi wrote:
| Legal rights are signed away all the time in contracts
| though.
| margalabargala wrote:
| The TOS specify the circumstances in which the corp may
| take action that is unrelated to the legal system. Just
| because they can't sue you (and easily win) for scraping,
| doesn't mean they can't block you if they notice you doing
| it.
|
| Google for example has a TOS and is well known for
| permanently banning accounts for real or imagined or AI-
| generated violations of it. Google banning you for breaking
| TOS doesn't mean you broke the law, just that you broke
| their rules, which apparently include a clause against
| being in the wrong place at the wrong time.
| SahAssar wrote:
| From that article:
|
| > However, after further appeal in another court, hiQ was
| found to be in breach of LinkedIn's terms, and there was a
| settlement.
|
| So why would the same not apply here?
| hrimfaxi wrote:
| They settled out of court, that doesn't mean that they were
| found to be in breach of the terms.
|
| These were some of the notable elements (worth noting that
| none mention breaching terms of service):
|
| > Damages: Judgment in the amount of $500,000 is entered
| against hiQ, with all other monetary relief waived.
|
| > CFAA liability: hiQ stipulates that LinkedIn experienced
| losses sufficient to, and "may establish liability" under a
| CFAA civil claim "based on hiQ's data collection practices
| and based on hiQ's direct access to password-protected
| pages on LinkedIn's platforms using fake accounts."
|
| > California "CFAA": hiQ stipulates that LinkedIn "may
| establish civil liability" under California's state-law
| counterpart to the CFAA based on hiQ's data collection
| practices, use of fake accounts and other means to evade
| detection by LinkedIn, hiQ's direct access to password-
| protected pages on LinkedIn's platforms using fake
| accounts, and hiQ's unauthorized commercial use of data.
|
| > Trespass: hiQ stipulates that LinkedIn has established
| judgment as to liability under California law for the
| common law torts of trespass to chattels and
| misappropriation.
|
| > Irreparable harm: hiQ stipulates that LinkedIn has
| established that it has suffered an irreparable injury and
| that LinkedIn satisfied the remaining factors and is
| entitled to a permanent injunction.
|
| https://natlawreview.com/article/hiq-and-linkedin-reach-
| prop...
| exe34 wrote:
| it's only legal if you have a team of lawyers though. the law
| still applies to the rest of us.
| paulnpace wrote:
| It is the future: I own nothing, and I've never been
| happier. They can sue me and take nothing.
| exe34 wrote:
| I've been trying to convince myself I'd be able to live
| like Diogenes, sleep in the streets, bathe in the sea and
| just generally survive off scraps - but I think that only
| works if other people can afford to throw away scraps.
| portaouflop wrote:
| If you live in the west it's no problem - the amount of
| waste there is insane
| exe34 wrote:
| It won't be after the billionaires own everything and the
| rest are living off scraps already.
| 6stringmerc wrote:
| Tell that to judyrecords with the same smug attitude.
|
| Your textbook versus reality conceptualization of things is
| dogshit. It's exploitation to do what OP did. You're
| endorsing it and minimizing the ethics and this certainly
| shall poison the well from which you drink. Godspeed.
| kulahan wrote:
| This is so overly dramatic it's hard to even consider the
| point you're trying to make.
| hrimfaxi wrote:
| > CFAA liability: hiQ stipulates that LinkedIn experienced
| losses sufficient to, and "may establish liability" under a
| CFAA civil claim "based on hiQ's data collection practices
| and based on hiQ's direct access to password-protected pages
| on LinkedIn's platforms using fake accounts."
|
| This was part of the terms of the settlement.
| pantropy wrote:
| Technically speaking none of Goodreads material or content is
| being used publically, the only information displayed on the
| site is freely available (Title, Author) and not Goodread's
| property.
|
| You could try to argue that this falls under "create derivative
| works from any materials or content accessible on the Service"
| but even then it seems really flimsy to say that recommending
| books based on Goodread reviews is an infringemnt.
|
| It's just not that different to a youtuber saying "I read
| reviews for 50 books, here's the ones to read"
| voidUpdate wrote:
| I'd be impressed if a youtuber could read 3 billion reviews
| and recommend books to you based on that
| tonyhart7 wrote:
| what about youtuber that build a machine that scrape 3
| billions books and make recommendation based on the
| data????
| bravoetch wrote:
| Skip that step. This project enables a Youtuber that
| automates pulling related booklists from this site, and
| uses AI to make the recommendation videos. Thousands of
| videos.
| croes wrote:
| I visit your garden and take 1 apple from your tree
|
| I visit your garden and take 1000 apples from your tree.
|
| Not that different.
| kemotep wrote:
| Not only am I taking 1,000 apples, but I use those 1,000
| apples to start my own orchard and encourage people to come
| to it instead of yours.
| simianparrot wrote:
| Yeah but if I program a drone swarm to automate this
| process it's for the greater good -- more apples for
| everyone!
|
| And I only charge a tiny subscription for access to all
| my drone-managed orchards, you can eat as many apples as
| you want. But don't steal any and start your own orchard
| or I sue.
| croes wrote:
| All the people who care for the trees and pick the apples
| have lost their job while an apple became nearly
| worthless, but without a job it's still unaffordable.
|
| Replace your drones with China or India and you have the
| current situation in the US.
|
| Apple farmers go out of business so you lose the people
| who create new varieties.
| mejutoco wrote:
| ... and somehow your garden did not lose any apple in the
| process.
| croes wrote:
| But you are an apple seller
| lm28469 wrote:
| > but I use those 1,000 apples to start my own orchard
|
| Steal cuttings, not the fruit, if you plan to start an
| orchard. From 1000 apples you'll get ~10 000 seeds,
| statistically you won't even end up with one good tree.
|
| > An output of three cultivars from around 50.000 seeds
| means that 17.000 seeds were needed to get one cultivar.
| Only one out of around 9.000 scab resistant seedlings
| showed the appropriate quality to become a cultivar. This
| proportion underlines the enormous effort which is
| necessary to develop a new cultivar.
|
| https://orgprints.org/id/eprint/13698/1/220-225.pdf
| jalk wrote:
| Not a great analogy, since a digital copy leaves the
| original intact unlike your apples
| MichaelBosworth wrote:
| What expectation of confidentiality are you ascribing to people
| having posted publicly accessible opinions on the internet?
|
| Out of curiosity, is your point about TOS out of concern for
| the poster or for Goodreads?
| voidUpdate wrote:
| My expectation isn't of confidentiality, but of attribution.
| Sure, my website is perfectly accessible on the internet, and
| I'm fine with being able to find it on google, but if you
| pipe it into an algorithm that will start throwing out stuff
| based on what I wrote, with zero reference to me at all, I'd
| get a bit annoyed. This website has taken the combined output
| of probably thousands of people, shoved it into an algorithm
| and is then using their work to give "original" ideas. If one
| person wanted their content removed from the system, how
| would you do that?
| caconym_ wrote:
| What does that comment have to do with confidentiality?
| MichaelBosworth wrote:
| That he viewed a review on Goodreads as the reviewer's
| intellectual property hadn't occurred to me. I see why, in
| aggregate, many such opinions become valuable, but the
| whole is more than the sum of its parts.
|
| So does it feel to you guys like your comments, say, here
| in this Hacker News thread should be considered effectively
| copyrighted as your personal IP?
|
| If so, do you feel the same way about opinions you share
| out in a supermarket or on the street?
| caconym_ wrote:
| There are well established legal standards for what is
| copyrightable and I believe written literary criticism
| trivially qualifies (as it should). Stuff you yell at the
| supermarket doesn't, IIUC, as it isn't fixed in a
| tangible form. Social media comments are, IIUC, generally
| protected. The exception would be comments that don't
| meet the bar to be considered "original", "creative",
| etc.
|
| (not a lawyer)
| lunias wrote:
| If it's on the internet, and people can access it, then it's
| public. I would have no expectations for what people do with
| public data; that just seems like setting yourself up for
| disappointment.
| Vvector wrote:
| Is a pirated movie, found on bittorrent, public?
|
| IMO, your definition is overbroad
| lunias wrote:
| If it's on bittorrent then, yes, it's public. It doesn't
| matter if you intended it to be or not, it's publicly
| accessible, therefore it's public.
| contravariant wrote:
| At what point are they feeding reviews into an LLM? From what I
| got the only personal data they're using is which user read
| which books.
| irl_zebra wrote:
| This is, essentially, why I've withdrawn from posting content
| from my human brain almost anywhere on the open internet
| (except here, sometimes) and have retired blog posts, opinions,
| and so on to our friends WAN.
| laszlojamf wrote:
| I gave it a spin with Gravity's Rainbow. Most of the
| recommendations are what you'd expect, Pynchon himself, Don
| Delillo, David Foster Wallace... and then right at the end... The
| Hobbit, or There and Back Again >.<
| tgv wrote:
| It seems you cannot reject recommendations. Based on two books,
| the system showed reasonable recommendations, some of which I'd
| read, including one that I really didn't like. There must be
| useful information in that, too.
| Llamamoe wrote:
| My impressions are similar as those of others:
|
| - It seems to mostly show me books I've already read and know of,
| including sequels of what I added, which isn't very useful.
|
| - It ultimately seems to prioritize "highest rated in category"
| too much, rather than focusing more on what made my chosen books
| stand out over others.
|
| - Needs a "disliked books" list, especially when the
| recommendations show me a lot of superficially similar books I
| hated. I'd like to blacklist them.
|
| - Would be cool to have a discovery mechanism for less popular
| and even obscure titles. Again, the top picks of each category
| are very well-known.
|
| - Might not be practical, but I'd like some way to filter by
| specific features of reviews. E.g. prioritize reviews that say
| "the MC is a psychopath/murderhobo/rapist" higher for anti-
| recommendation, ignore reviews that say "whiny character", etc.
| mna_ wrote:
| I typed "Introduction to Real Analysis by Bartle" and I got:
|
| Steve Jobs by Walter Isaacson
|
| Harry Potter and the Sorcerer's Stone (Harry Potter, #1) by JK
| Rowling
|
| Topology by James R Munkres
|
| and so on.. Munkres' book is relevant and I want to read it, but
| what have Steve Jobs and Harry Potter got to do with with
| mathematics?
| franticgecko3 wrote:
| They have nothing to do with mathematics but everything to do
| with being extremely popular books.
|
| Most people that have read a mathematics textbook have also
| read and enjoyed Harry Potter.
|
| Given you have enjoyed drinking water and breathing in the
| past, there is a high likelihood that you will enjoy watching
| the Star Wars films.
| costco wrote:
| You may find these better:
|
| https://book.sv/similar?id=211570
|
| > Note 1: If you only provide one or two books, the model
| doesn't have a lot to work with and may include a handful of
| somewhat unrelated popular books in the results. If you want
| recommendations based on just one book, click the "Similar"
| button next to the book after adding it to the input book list
| on the recommendations page.
| atomicnature wrote:
| Will this be open source?
| sexylibrarian wrote:
| We've been working on this data set since 2016 and have it
| covered! Our app is on test flight in private beta and will be
| sharing it very soon xo
| caro_kann wrote:
| I tried three different genre books, for each I got 1984 - George
| Orwell recommendation.
| diffeomorphism wrote:
| The robots.txt is pretty explicit that this scraping is
| "disallowed"
|
| https://www.goodreads.com/robots.txt
|
| So legalities aside, this seems unethical.
| galdauts wrote:
| I agree. As a frequent reviewer on Goodreads, this feels really
| icky.
| psandor wrote:
| You are right.
|
| At the same time, everything you ever posted online has
| already been scraped by hundreds (maybe thousands) of
| entities and distributed/sold to countless other entities.
| The only difference is that OP shared his project here.
| paulnpace wrote:
| It is truly criminal that such a bright and brilliant model of
| ethics, Amazon, should endure such an attack.
| diffeomorphism wrote:
| Unethical behavior does not become good just because it
| happens to hurt "bad people" (or more accurately, companies
| bought by bad people).
| paulnpace wrote:
| Using a sword to stab someone is evil, therefore, stabbing
| someone who is stabbing me with a sword is evil?
| wongarsu wrote:
| Another factor is that Amazon is big enough that crawling
| a minor website under their umbrella for a noncommercial
| project is unlikely to notably affect them.
|
| Stabbing people with swords is evil, unless they are so
| big that to them it's at worst a light poke with a fork
| sputr wrote:
| Why would it be unethical?
|
| This obsession with "everything must be commercialized" is
| really killing creativity.
|
| Now if the author was commercializing other peoples reviews,
| sure, it's potentially(!) unethical. But scraping a website for
| reviews that are publicly(!) posted, training a recommendation
| LLM and then sharing it, for free, seems ... exactly the ideal
| use case for this technology.
| contravariant wrote:
| If it's unethical it's not because of what the robots.txt says.
|
| _Blindly_ violating it is bad manners, but deliberately
| scraping a single website over a month isn 't the worst.
| billfruit wrote:
| I feel that the last added book in one's list seem to have more
| influence on the recommendations, which results in a rather
| similar type of recommendations.
| costco wrote:
| This is a result of the use of positional embeddings, which
| typically results in the final item being weighted very highly.
| The problem is that this information is shown to be very
| relevant to the task of predicting the next item interacted
| with. If you add more books the effect of this is somewhat
| diluted.
| logicprog wrote:
| I tested the model out, and I can attest that it's very on point!
| It gets the assignment
| fridental wrote:
| I've entered books from The Expanse and Lockwood & Co series and
| its output was not really overwhelming: - other books from the
| series (duh, I don't need a recommender for that recommendation)
| - Hobbit, Harry Potter, Azimov etc (duh, I like scifi and surely
| I've already read all the classic works).
| conartist6 wrote:
| SHAME. Gross. Morally bankrupt. Greedy.
| stichers wrote:
| I don't get what "Add" does
| hasbot wrote:
| Try it and see! But, it adds the recommended book to the list
| of selected books.
| deanc wrote:
| Looks cool, and no bullshit. Please let us filter
| recommendations. If I put in a non-fiction book I'm probably
| looking for recommendations of other non-fiction books :)
| boplicity wrote:
| Please remove my reviews from your LLM model and training data.
| Thank you.
| mparnisari wrote:
| Why?
| smcleod wrote:
| Very neat! By chance have you open sourced the code and model
| anywhere? I'd be really keen to have a play with this.
| piskov wrote:
| Tried and unfortunately it was meh.
| mhb wrote:
| Import doesn't work for me.
| runnr_az wrote:
| same. will stop back by in a day or two when the site has a
| chance to recover
| aidenn0 wrote:
| I have "too many books" to add more to books I like and I've read
| almost all of the recommendations...
| 6stringmerc wrote:
| Wow I wonder how fun conversations in the afterlife with Aaron
| are going to be for the OP. There are ways to improve broken
| systems. Pimping them out with glee is not one of them in my
| book.
|
| Everything about this concept I hate and it's difficult not to
| conflate that with the creator. I make comparisons and
| equivocations. This is an ethical discussion akin to "Can you
| enjoy Bill Cosby comedy knowing he was a Rapist" and I'm not
| being glib.
| Svoka wrote:
| Honestly, with this I see same results as with any other
| recommendation system - I type some nice Sci-fi/Fantasy I read,
| it give me generic Sci-Fi fantasy I already read. Even those I
| really didn't like.
|
| I add those I like to the list, ignoring those I didn't, and in
| the end I just end up with recommendations I already read and
| didn't like.
|
| I feel like wasted my time yet with another smart recommendation
| system.
| greenie_beans wrote:
| so sick!! always wanted to do something like this but would use
| it for my saas (https://bookhead.net), so i'm hesitant to scrape
| goodreads due to their terms of service.
|
| in addition to the goodreads data, i think it would be
| interesting to add an author's favorite books by scraping their
| paris review interviews (and other interviews) and using that as
| a "review" because i've learned about so much good stuff through
| an artist mention.
|
| and in a retail context, i've always wondered if a recommendation
| engine like this could have its own "flavor" based on a specific
| store's customer buying history. like if a bookstore's customers
| were weighted in the algorithm so that their similarities scores
| were given preference. much of what a bookstore carries is based
| on their customers' taste. you could use the goodreads etc as the
| base recommendations and then train it on a bookstore's sales
| history.
|
| a project like this is a bit outside of my expertise but i have a
| tiny bit of knowledge about it, and now i have a lot more to
| learn after seeing this. if anybody has any good book
| recommendations (hehe) or papers i should read to learn about
| these sort of systems, please let me know!
|
| thank you for sharing!!
| Hilift wrote:
| This seems like it should be an easy task for an AI to implement.
| For example, the question "what is the most helpful rated
| negative review of the book 'Original Sin' by Jake Tapper?" There
| are obvious and prominent "helpfulness" ratings of reviews, but
| they don't seem to be scraped, at least not by Gemini.
| Additionally, Gemini reports seemingly inaccurate or minimal
| effort information:
|
| "It is difficult to pinpoint a single "most helpful" negative
| review of Original Sin by Jake Tapper, as helpfulness ratings on
| platforms like Amazon or Goodreads are dynamic and subjective,
| and the provided search results *do not include specific user
| reviews with their respective helpfulness votes*."
|
| https://share.google/aimode/OnWrGe4j508c4u3gh
| Invictus0 wrote:
| The problem with recommender engines is they're always
| recommending the most popular books that are in the same vein as
| what you've already read. So you're always getting pop-culture
| pap and not actually-interesting, somewhat more niche and
| unrelated books that are only tangentially related. The
| recommendations I got were all pop-psych stuff and other titles
| by the authors I've already read.
| aaronax wrote:
| Can you create a list of the most common book names? I think it
| would be funny to have a shelf of books that all have the same
| name.
| mring33621 wrote:
| FEEDBACK:
|
| I should be able to mark recommended books as "Read, Liked";
| "Read, Didn't Like"; "Remove, Other Reason"
|
| and then allow a rerun of Recs, based on additional info
| calebt3141 wrote:
| This is fantastic. Thanks for sharing this.
| fsmv wrote:
| Cool so you stole my data and now you're bragging about it?
| mayahisali wrote:
| Curious about how you trained the model on a billion reviews.
| What architecture did you use?
| contravariant wrote:
| For the intersect page you probably want to order the users by
| the size of their shelf. For some more obscure combinations I'm
| mostly getting users who read 10,000s of books, which is less
| useful than the users with <1000 books.
| sosuke wrote:
| Feature request: Combine book series into a single entity. Bummer
| getting recommendations for another book in the same series as
| one I already liked and read.
|
| Feature request: Exclude books already in shelf. This is harder
| I'm sure. I've got 1146 books in my Read shelf.
| tecleandor wrote:
| _sigh_ No thanks. I rather not have my comments used by random
| LLMs.
| hmokiguess wrote:
| People of HN: - Remove my data, unethical, my words are mine!
| Argh!
|
| Also People of HN: - I built an HN aggregator that shows
| sentiment analysis of comments and . . .
| freen wrote:
| The commons of the creative output of humanity is a resource,
| just like oil or lithium.
|
| We are rapidly replaying the worst of the Resource Curse.
| bossyTeacher wrote:
| You scrapping Goodreads to make a replacement is like your
| company making you train the new hire so he can replace you
| eventually
| wtf242 wrote:
| That's super cool. I launched a book recommendations feature this
| year, which works vastly different. I ask users what their
| favorite books are(which you can rank), and then allow them to
| import their goodreads data which includes star reviews, and
| books they have read, then I determine your favorite style of
| books based on genres and subjects, then use opensearch to find
| similar books. It's a lot more complicated than that, but seems
| to work well. I'm always looking for ideas on how to improve this
| feature. Interested in what you are actually doing on the backend
| on the how-it-works page. thanks!
|
| here's my recommendations feature:
| https://thegreatestbooks.org/recommendations
|
| It's much more powerful if you're a member. you can restrict the
| results to certain genres or book lengths, as we as published
| date ranges, etc. If someone wants to try out the more powerful
| feature, DM me and i'll mark your account as a member.
| djent wrote:
| Why not scrape the content of the books too?
| loremm wrote:
| For intersect I also wonder if you add a filter that the books
| are within the top rated. Like if I give my favorite books and
| want to find someone who has my same taste, it doesn't help if
| they hated (all/most/some) of those books. Tricky in that not all
| users give star ratings
| mhb wrote:
| Do you have any thoughts on how SASRec compares with the SVD-
| based Cinematch algorithm?
___________________________________________________________________
(page generated 2025-11-07 23:01 UTC)