This is a guest post from Leila Roy. Leila has a regular weekly column about YA fiction at Kirkus Reviews, and started Bookshelves of Doom way back in 2004. In the physical realm, she’s a library director in rural Maine and watches a shocking amount of television. Follow her on Twitter @bkshelvesofdoom.
I’m a librarian and a book reviewer. At any given time, I have approximately eleventy billion books to choose from. And yet, sometimes I still feel like I don’t have anything to read. Either that, or I just don’t want to read any of the books sitting in front of me.
Enter book recommendation engines! Websites that answer that most difficult of questions: WHAT TO READ NEXT?
But that leads to yet another decision to be made: WHICH ONE TO USE? There are a whole lot of them out there, and I’ve done a lot of digging around and a lot of testing. Based on pure reliability—as in, the website is up more often than not, and provides recommendations more often than error messages—I’ve narrowed the best options down to four. But even then, which of those to use? To answer that question, I give you: THE BOOK RECOMMENDATION ENGINE CAGE MATCH.
In Corner #1, we have Your Next Read, the self-learning engine that relies on user ratings and suggestions to build an image-based map of recommendations.
In Corner #2, we have What Should I Read Next, the text-based contender that uses user-entered booklists to create connections between seemingly disparate titles.
In Corner #3, we have NoveList, the database for fans of Boolean operators that matches books by subject heading, tone, and genre.
And finally, in Corner #4, we have Goodreads, the website best known for being a hotbed of literary snark and GIF-heavy book reviews, but which also provides personalized recommendations based on “20 billion data points.”
Round one: Visuals
NoveList and Goodreads both offer up a TON of information on each page, but that much information can make navigation confusing for new users and somewhat irritating for regular users who’re simply looking for the answer to a very specific question.
What Should I Read Next, which doesn’t provide any images at all, is on the Zork end of the spectrum.
Your Next Read, meanwhile, uses an intuitive click-and-slide interface that also provides an ongoing visualization of your search history.
Winner: Your Next Read
Round two: Ease of Use and Accessibility
All four sites allow for simple title, author, and ISBN searches right from the home page, though What Should I Read Next makes you jump through an extra hoop before accessing the ISBN function.
American public libraries generally offer free access to NoveList, but the fact that it’s only accessible with a username and password is a huge strike against it in this category. And while it’s certainly usable by non-librarian folk, it’s also very much geared towards the library world, in terms of the advanced search functions and the jargon and even the assumed audience of the articles (librarians). The other three services are usable without signing in, though all three also offer the option to create a free personal account.
Winners: Your Next Read and Goodreads
Round three: Other Features
Your Next Read provides some pre-made booklists… but the featured lists are mostly from 2010 and 2011. NoveList has lists of recommendations based around current bestsellers, as well as librarian-created articles about author read-alikes, suggestions for book-to-movie picks, and the capability to search by publication date, suggested age range, the author’s cultural identity, and book length. GoodReads is more of a social media site, so in addition to the community element, it offers the most of all, with user-created booklists, author events, giveaways, and book clubs.
Winners: NoveList and Goodreads
Final round: Recommendations
And now we come to the final and most important determination: which sites actually give reliably good recommendations? To test them, I ran a selection of titles through each engine: The Greenglass House, by Kate Milford; Smile, by Raina Telgemeier; Gabi, a Girl in Pieces, by Isabel Quintero; 1st to Die, by James Patterson; Devil in the White City, by Erik Larson; Fingersmith, by Sarah Waters.
To the chart (split it two for readability)!
A few more notes:
*Your Next Read, especially, offers up a lot of titles by the same author—a practice I view as weaksauce, as most readers are likely to have already done at least that amount of legwork on their own.
*If you’re looking for comic book recommendations, neither Your Next Read nor What Should I Read Next will be particularly useful—I tried a number of other titles, and the recommendations from Your Next Read were less-than inspired, and most weren’t indexed at all at What Should I Read Next.
*The Novelist recommendations are compiled from a combination of books specifically recommended by librarians, and books that, according to the computer brain, have elements in common with the original title.
*Recommendations are tricky. Three different readers could ask for a book “like Fingersmith” and be looking for three entirely different things—a period drama, a mystery steeped in economic class, a lesbian love story—which is why none of these sites can compete with a good bookseller or librarian.
I admit it, I was rooting for one of the two scrappy upstarts to carry the day, but it can’t be denied that the winners, this last round and overall, hands down and no question, are Goodreads and NoveList. Overall, while Your Next Read and What Should I Read Next are entertaining diversions—though neither holds a candle to the sheer geeky joy that the Tourist Map of Literature brings me or even to the entertainment value I get from the sliders at Which Book—Goodreads and NoveList offer stronger, more trustworthy recommendations for a far broader selection of books.
What about you? Do you use any of these sites? Or have you discovered a different favorite?
Book Riot Live is coming! Join us for a two-day event full of books, authors, and an all around good time. It’s the convention for book lovers that we’ve always wanted to attend. So we are doing it ourselves.