This post is the third part in a series aimed at helping readers to more effectively identify books they will enjoy—that is, to invest their reading time wisely, and to reap all of the dividends accruing therefrom.
If you Google “book recommendation engines,” you get a Google-typical 26.9 million results (in 0.25 seconds, no less).
The very first result in that list is the promisingly titled WhatShouldIReadNext.com.
The results of my test search, however, were much less promising. When I told it I had read and enjoyed Marilynne Robinson’s Gilead, these were the first few suggestions:
- Return of the Prodigal Son by Henri Nouwen (a book about “spiritual life”—nonfiction, so far as I can tell)
- An Unfinished Season by Ward Just (according to the site, a psychology book, though Amazon assures me it’s a novel)
- Crooked Little Heart by Anne Lamott (definitively a novel)
- The Art of the Infinite by Robert and Ellen Kaplan (subtitle: “The Pleasure of Mathematics”)
…and so on. That list strikes me as no more immediately useful than Amazon’s tabulated sales data (as discussed in the last post in this series).
Another engine, BookSeer.com, borrows the conceit of the old AskJeeves.com but, disappointingly, simply quotes Amazon’s “recommendations” verbatim. (I’m not kidding or exaggerating, though I very much wish I were. The list is titled “Amazon recommends:” and then reports exactly what you’d see on the corresponding Amazon page.)
When I mentioned to a friend that I was writing this article, she touted a newly launched site called Bookish.com. I tried it out in the interest of due diligence. When I typed in “Gilead,” it told me: “Sorry, no recommendations available for that book.” Hmmm.
A startup site called WhichBook.com shows significant promise by borrowing the Pandora approach (also discussed in the last post): they categorize books on the basis of twelve different dimensions, including happy vs. sad, funny vs. serious, short vs. long, and easy vs. demanding. If you specify your preferences, then the site will scour its database to make you an appropriate recommendation.
WhichBook’s approach is intriguing. The downfall, of course, is that when I told Pandora I liked the Zac Brown Band, it inferred that I must like the features it had already associated with the group and then made additional recommendations based on those features. What it didn’t do was ask me to specify on a sliding scale just how much acoustic sonority I like in a song.
For the same reason, I wouldn’t even know where to begin with the funny vs. serious sliding scale. Why can’t I have both?
Pandora is a mature system that saves users the trouble of dealing with interlevels. It does the work for you. Users provide input by telling the system what songs they like and receive output in the form of other songs they will probably like. They don’t have to worry about why.
WhichBook, in turn, aspires to Pandora’s level of user-friendliness, and it may well reach it. Unfortunately, it’s not there yet, and the amount of data that will have to be crunched to get there is no doubt overwhelming. Still, the site bears watching.
Another site with promise is BookLamp.org, which is part of the so-called Book Genome Project. (For comparison, Pandora operates on the strength of the Music Genome Project.) The site is not currently making recommendations, but there’s a lot of DNA left to be mapped. I set a Google Alert years ago in hopes that it would eventually revolutionize the book-recommendation game.
In short, the existing options for book recommendation engines show promise but still fall short. The blog Bookshelves of Doom compiled an exhaustive survey of a number of similar sites, none of which I found overly impressive.
If you, dear reader, know of other sites that you feel warrant inclusion on this list, then please tweet BMR @BlueMesaReview to let us know. Given sufficient interest, I’ll write a follow-up post to detail all of your recommendations.
The most popular book recommendation sites rely more heavily on the strength of their community than on any particular algorithm. Sites like Shelfari, GoodReads, and LibraryThing strike me as nothing quite so much as book-based social networking, Facebook for the literati. All three are, in both my prior experience and my rigorous scientific research in advance of writing this article, excellent compendiums of reviews. But they all seem to require too much effort.
I don’t want to invest the time to create a parallel Facebook profile. I don’t want or need a virtual bookshelf. I don’t have any interest in joining an Internet-based book club. I just want a stinkin’ book recommendation. All three sites strike me as incredibly inefficient means to that end—especially when you consider that the user rating-driven “recommendations” they make are often very far afield from anything you’ve actually rated.
Moreover, I don’t really have any issue with Amazon reviews.
The problem with any review is, of course, whom to trust. If I read your GoodReads review of Gilead, how do I know that your taste in books is anything like mine? To wit, I just read Jennifer Egan’s A Visit from the Goon Squad at the insistent recommendation of several readers I hold in very high regard—and I hated it. I honestly have no idea why it won the Pulitzer (though I suppose I could make a PowerPoint to speculate as to the reasons).
BookSeer.com notes the following beneath its reproduction of Amazon’s “recommendations”: “Of course, you could go ask your local bookshop or your local library.”
That’s the premise of sites like TheStaffRecommends.com, which not only suffer from the impersonality of the Internet, but also bear the taint of naked commercialism. I don’t want to wonder why one book is being pushed over another. I just want a book recommendation from a source I can trust.
Fortunately, I have just such a source. I was lucky enough to stumble across the world’s best book recommender some four years ago now, and that source has yet to let me down. In fact, that’s where I found out about Marilynne Robinson in the first place.
Next week I’ll reveal my secret weapon in the search for great books, an ongoing struggle that has claimed far too many good readers, but one that need not lay waste to any more of your precious reading time (a nonrenewable resource if ever there was one). BMR may not be able to slow down the quicksand of time, but we’ll do what we can to help you spend it the way you want to. Check back next week for the lifeline that can pull you out of the morass.
Michael Noltemeyer is a third-year MFA candidate at the University of New Mexico. He is the Nonfiction Editor for Blue Mesa Review.