A recommendation engine informed by gestures

[UPDATE, June 13, 2011: We’ve activated the eleet.fi domain and blog.]

(Cross-posted on the Uutisraivaaja blog)

We consider the above tweet by New York University Journalism Professor Jay Rosen to be the most flattering endorsement to date for the idea of ‘Gestures’, or ‘Eleet’ in Finnish.

Our vision is to create a personal recommendation engine informed by on-line social gestures.

Wake up your favorite glowing rectangle and your Internet browser, RSS reader, Facebook or Twitter client, or special-purpose mobile application will show you hyper-relevant Internet destinations.

The way in which we interact with on-line information is informative of our news preferences.

Find me stuff I’m interested in.”
Dave Winer

Subscribing to an RSS feed; reading, storing, sharing, tagging, rating or sending an article; writing a blog post, commenting on or linking to one: all these actions can be seen as ‘gestures’, indicating that the user attributes some degree of relevance to the content at hand.

By analyzing the gestures expressed by different people, we can infer who have similar preferences for news, and then offer news of similar topics to people with matching interests.

We intend to launch the first public deployment of our recommendation engine during the summer of 2011 with a minimalist feature set. Then release often with incremental improvements and extending the features.

Given Facebook’s significant adoption in Finland, our first release will most likely offer Facebook authentication, and the first recommendations will be based on links people share on Facebook. In other words, those Facebook links will be the first gestures informing our engine.

Our team is small but enthuzed. During a kick-off workshop in May, we identified various technical components to work on – see above flip-chart image 🙂

One frequently asked question by people who listen to our idea is about echo chamber syndrome. We realize that people want surprises in their news diet. Therefor, to address this concern, we intend to build into the UI a way to vary the level of serendipity in a user’s recommended content.

An interesting piece of feedback we received at the Uutisraivaaja pitching event in Helsinki on May 12 was about positioning this service.

For sure there are many recommendation systems around. Of course we believe that the quality of our engine will be different because of the concept of gestures. But still, the point of positioning is valid.

[T]hinking about information overload isn’t actually describing the problem – and thinking about filter failure is.”
Clay Shirky

So, one idea that came up during the pitching event was to position this recommendation service as offering something different, something to let it stand out from the pack.

For example, it was suggested, instead of serving the most relevant, personalized news headlines to a user, how about serving news which is preferred by someone whose   profile is quite the opposite of the current user?

Interesting… we may be able to do that, if it’s what users want.

Short-story-long: look out for eleet.fi in the course of June. If reading this gets you all wound up and excited, why not join our little group and help create some awesomeness? 🙂


4 thoughts on “A recommendation engine informed by gestures

  1. eleet.fi returns nothing yet, hope you registered it,though…

    Cool idea btw, seeing the benefit of it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s