• Print

Towards a better book recommendation service

The ideal discovery platform requires not one, but many input sources

The ideal content discovery service has yet to be invented. Plenty have tried but none have truly succeeded. The latest is venture is BookScout from Random House. It’s a nifty Facebook app that uses your social graph to help you discover relevant content. As Laura Hazard Owen recently discovered though, it’s far from perfect.

Reading Laura’s post reminded me of something a wise person told me last year: Just because I’m Facebook friends with you doesn’t mean we have the same reading interests. In fact, I’d be willing to bet my reading interests don’t map very well to any of my friends, real or virtual.

That’s the problem. We try to take one aspect of our lives and have it spit out book recommendations. That model is doomed to fail every time.

I’d like to propose a completely new book recommendation model. Rather than just looking at my Facebook friends, who I follow on Twitter or any other single activity, why not roll them all together and build an algorithm around everything I do online?

Monitor my Gmail. Track every website I visit. Keep a record of the various searches I do on Google. Log all the books I look at on bn.com, Google Play and every other catalog I visit. Keep a close eye on the RSS feeds I actually read. Study the Facebook posts I comment on and every word in each of my tweets. Look at everything I clip on Findings. Use it all!

Privacy freaks are ready to explode at this point. They can’t imagine why anyone would allow themselves to be tracked this closely. Fine. They don’t have to participate. I would though, especially if it leads to better recommendations. And I bet it would.

tags: , ,
  • http://profiles.google.com/edward.w.bear Edward Bear

    Major distributors (yes I mean YOU Amazon and B&N) can improve things quite a bit for authors you already know by adding a “Notify me about new books by this author” button on each book page. Fictionwise had a service like that ten years ago and a huge chunk of “I know this writer” authors on my reading list benefited from it. The current situation of having to check (repeatedly) multiple websites and do searches for 30-50 authors is simply untenable.

    New discoveries, OTOH do need better mechanisms too, but I don’t regard them as quite as critical, since I’m already drowning in reading material.

  • John E Rollason

    Or you could ask friends for personal recommendations?
    Radical I know but might catch on.

  • http://profile.yahoo.com/TBZ5A2S756QZG5FPKZDQU2KUWM DPEvers

    Perhaps try one of the many, many on line reading list sites: enter book read, enter rating, receive recommendations.

    Other than book searches and reading book reviews, I think little of what I do on line would be a good predictor of my reading tastes.  Even reading of book reviews isn’t a good predictor, because I end up not liking the book after I read a review or two.

  • http://about.me/mvagnetti Michael Vagnetti

    Great post. Call it “context mining.”

    I’m fascinated with the connection between discovery and how we actually write. We often talk about “the reader” or “the consumer.” How does a reader’s/consumer’s writing have the potential to take discovery from “OK” to “amazing”?

    Think of markets as conversations, and of “books” as as the variety of contexts (web or print or both) around which those conversations happen. Buying, reading, and writing are all part of the same activity: creative capital.
    A great “discovery engine” would be focused on giving people “an amazing conversation” by encouraging all of those activities. The writing piece seems to be the elusive part, but powerful. (For example, great reviews – Amazon, Goodreads – make for more useful discovery.)

  • William Ockham

    Better book recommendations won’t come from your obeisance to Big Brother Google. Better book recommendations will come from knowing more about the books you like. The patterns that need to be recognized are in the books, not in your meanderings on the web.

  • Janet

    NoveList. Been around for decades at your local public library–which you can access online if you don’t want to get up and go where the books are.

  • http://twitter.com/brandizzi Brandizzi

    I wonder if this is not a useless goal by itself. Really, don’t you people already have more books to read than one can accomplish? Do you really need a service to  make your stack still bigger? If so, why not leaving it to the real life, for the fate, for your friends? Maybe this is the reason no such service catches on – nobody needs it.

    • http://www.georgebaier.com/ George Baier IV

      I have 385 on my goodreads ‘to read list’… but I MUST ADD MORE.

    • http://profiles.google.com/edward.w.bear Edward Bear

       The eight most terrifying words in English are: “I have run out of things to read.”  Overstock, toujours l’overstock!