For Your Reading Pleasure

If you’re a bibliophile, nothing beats going into your favorite bookseller and having a clerk who knows your reading habits suggest a tome or two that you’re sure to love. Now you don’t even have to leave home to get such personalized help.

Cyber book purveyor Amazon.com launched a new software program in September that combs through its database, looks for people who have purchased the same books and comes up with a short list of suggestions that it offers repeat customers when they come back to its Web site.

“Hello mike ybarra,” beamed the friendly cyber-clerk when I stopped by the Amazon.com home page recently. “got recommendations just for you! (If you’re not mike ybarra, click here.)” Come back soon, the obliging site urged, as selections change daily.

“Customers absolutely love it,” said David Risher, a senior vice president at the Seattle-based company. “Everyone has in their past a book that has changed their life and if we can introduce someone to one of those books we’ve got a customer for life.”

The software (an in-house version of Net Perception) picks a few books in the same general area, throws in a couple of choices and screens out controversial or potentially offensive titles.

The other day, for example, a visitor to the site who had recently purchased a collection of John Dos Passos novels and a biography of journalists Joseph and Stewart Alsop signed on to find the following recommendations: a collection of John Muir’s writings, another Dos Passos book, the latest novels by Don DeLillo and Thomas Pynchon, and a science fiction book by an obscure writer. Except for the last selection, all the titles at least sounded interesting. “People will occasionally write and say, ‘I’m not sure why you recommended this,’ ” Risher said.

Still, even software has feelings. The last time I signed on looking forward to some personal recommendations, my cyber-clerk apparently thought I had been neglecting my end of the relationship. “Once you order some more books,” sniffed the Web site, withholding its suggestions like the romantic favors of a jilted lover, “come back to see our recommendations based on your purchases!”