addressalign-toparrow-leftarrow-rightbackbellblockcalendarcameraccwcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-checkcircle-with-crosscircle-with-pluscrossdots-three-verticaleditemptyheartexporteye-with-lineeyefacebookfolderfullheartglobegmailgooglegroupsimageimagesinstagramlinklocation-pinm-swarmSearchmailmessagesminusmoremuplabelShape 3 + Rectangle 1outlookpersonJoin Group on CardStartprice-ribbonImported LayersImported LayersImported Layersshieldstartickettrashtriangle-downtriangle-uptwitteruseryahoo

Ted Dunning: An Introduction to Polyvalent Recommendation

Speaker: Ted Dunning, Chief Application Architect at Mapr and Co-author of Mahout in Action

Topic:  A new approach to recommendation, polyvalent recommendation, that is both simpler and much more powerful than traditional approaches.  The idea is that you can combine user, item and content recommendations into a single query that you can implement using a very simple architecture.


Join or login to comment.

  • Timothy V.

    is sorry to have missed it. I missed meeting Ted and Paul O'Rorke, a former Sony Data and Statems Architect I just missed working with.

    As far as "Learning ot rank in a very bayesian way", I admit to being leery of a method of ranking that "...present(s) comments to a user, the metric would be sampled independently for each comment and the comments would be sorted according to the resulting scores."

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=955984

    ...why not present negatively rated comments with some threshold-override frequency just to make sure that the already over-self-sampling of a highly opinionated group doesn't self-reinforce to the point that it never even gets the chance to read an oppostional or alternative viewpoint!

    Exposure to only that content some algorithm decides you will agree with, "a priori", can be a potential "social danger", eh? ;^D

    May 17, 2013

  • Ellen F.

    This is an outstanding group - very friendly, great questions, good interactions. I look forward to seeing how it evolves with each new event.

    Thanks again to Netflix for providing the excellent meeting space and food, and to Will and Joel for organizing the event. Enjoyed meeting some new people tonight.

    For those interested in more from Ted, here's a recent blog post on a new approach to ranking with Bayesian bandits: "Surprise and Coincidence" http://bit.ly/13xwQIX

    Cheers, Ellen

    May 17, 2013

  • Will F.

    Ted gave an awesome presentation and Netflix was a fantastic host once again! Thanks to everyone who braved the traffic.

    May 16, 2013

  • Sree V.

    one great presentation from Ted.
    Commute was normal on 85 S / Winchester exit.

    May 16, 2013

  • Timothy V.

    discovered teh DunnigScore n LimeMinds personalization Server when I took over as Analytics Architec there, and have wanted to mee Ted ever since then.

    1 · May 13, 2013

  • Joel S H.

    Looks like I'll have to be on the road at the same time.

    May 8, 2013

  • Will F.

    We will increase the number of attendees when we finalize a location for this event.

    If you have a space that you would like to provide please let us know.

    thanks,
    Will

    April 28, 2013

    • Blake I.

      You're welcome again at Netflix if you'd like...

      April 28, 2013

Our Sponsors

People in this
Meetup are also in:

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy