Hmmm … A recommendation for books? Is it possible? And how would it work? I’m intrigued. The current website as is, appears to be something of a work in progress, serving as a means to demonstrate the technology and collect feedback. As with most things, potential success lies with how the user involvement goes. For now, BookLamp could be off to a promising start:
“Since the process requires BookLamp to scan the entire content of a book in order to catalog it, only the about 20,000 books from these publishers will be available at first. But unlike the current site, which is intended as a demonstration of the technology more than a consumer product, readers will be able to tweek their recommendation criteria. For instance, theyâ€™ll be able to request â€œsomething like the Da Vinci Codeâ€ but â€œless denseâ€ or â€œshorterâ€ or â€œmore fight scenes.â€
On the website: “BookLamp.org matches readers to books through an analysis of writing styles, similar to the way thatÂ Pandora.comÂ matches music lovers to new music.”Â
Following with that Pandora-like approach, what intrigues me the most was a description of a “Book Genome Project” and some thoughts on “Story DNA.” Interesting. Reading over it, one can’t help but feel a certain Roland Barthes-like approach to thinking about the contents of a book. I did a test run with The Girl with the Dragon Tattoo:
Their approach seems fresh and interesting. Of course, context is everything. Cool statistics are one thing, but we’ll have to wait and see how BookLamp goes with continuing to build its book catalog. I think we’ve all probably used Amazon’s book recommendations (“Customers who bought this item also bought …”), and BookLamp’s noncommercial alternative of a book recommendation engine could potentially very good thing.
In the meantime, here’s a videoÂ which provides a bit of a nuts and bolts look into what BookLamp is up to, and how.