While friends can give you book recommendations based on what they think you'll enjoy, sometimes they don't match the things you've actually read in the past. But now the New York Public Library is partnering with social bookstore startup Zola Books to give readers algorithm-based recommendations, which pull from book data rather than user data. Until now, the NYPL's recommendations have not been based on the reader's own activity, rather they've been based on what other readers have been checking out, viewing, and rating. The new service will give readers book choices to consider based on the characteristics of the books they search for.
Recommendations based on the book's personality, not yours
The technology comes from Bookish, a book discovery website that Zola recently acquired, which uses algorithms that rely on "deep, introspective" data to recommend similar books to the ones a reader searches for. The data it uses includes the authors, editors, and illustrators that contributed to a book, the genre and release date, and awards the books have won. Those recommendations will be built into the NYPL website and will appear on the side of the page of a book search result.
Currently Bookish can't offer personalized recommendations to readers like Netflix because it doesn't have enough user data to pull from. While there's no telling if these new recommendations will encourage users to rate and review books on the NYPL website, they could give readers more options to choose from when the book they originally searched for has already been checked out.