AlwaysOn first morning... “Beyond Search: Discovery”

First up is Tina Selig, giving the morning overview. Tina teaches here at Stanford. Smart and lovely woman! I don’t recall having any teachers like her.

The first panel is entiteld “Beyond Search: Discovery”

The suggestion is that we are about to get beyond the idea of purse search, and to move on to the era of discovery.

My style has been to try to get the best quotes from the speakers and try to attribute the ideas to the speaker, but things move fast so stay tuned.

Key idea is about consumer discovery on the web. Online equivalent of channel surfing. Need for discovery services comes out of what is happening in terms of digital content. Because there is so much, traditional search is breaking down. Trying to create targeted and relevant discovery.

The way these panels work, is that everyone gets a moment to make a brief pitch for what they do, and then the Q & A starts. 

Joe Greenstein Flixster… service that allows you to rate movies. Looking to see what their friends have said about a movie, really a form of sharing.

David Hyman, MOG, music discovery service, providing better tools for people to discover music. 1,500 music blog posts a week. Provice users with application called mogamatic, looks at your hard drive, your music, tool, allows you to see or show who you are! Use data for content filtration. Let you use this data set to filter the content. Find the people most like you musically.

Bambi Francisco Vator TV. Helping to aggregate information around ideas? Create networks around ideas. Relationships allow you discover new friends, ideas, etc.

Query, how does this stuff work under the hood? Through social networks? Build relevant recomendations? or Algorythmic approach…

Funny sitting here, thinking that you pay attention, and realize that I have never heard of any of these people or their sites.

Paul Martino….Aggregate Knowledge, founder of Tribe. Social based recommendations work, if you know the person making the recommendation. With only one kind of data set, you don’t get the magic that results that results from multiple contexts.

Greenstein….Discovering things through friends could be terrible, but could be an interesting recommendation. Don’t try to be accurate, try to be fun and interesting. Not focused on how accurate you can be. Part of the value of the recommendation is the fun of talking about it later.

Hyman… 450K music downloads from YouTube. Provides a form of discovery, technology exists for music identification. Track behavior, which is far superior to point of purchase data. The more you tell MOG about you, it can find other people who have similar music libraries. Magical as to how you are finding new bands, etc. Mog never recomends. Bad recommendations, leave bad taste.

Greenstein…What are you trying to discover, how big is the universe, how high is the cost of transaction. Discovering that I want to see a movie.

Hyman…Other cultural influences outside of friends, i.e. trusted sources that you don’t know. Example being that they are trying to bring on artists, and let you see what they listen to!

Francisco… Companies coming to Vator to see what people are doing.

Query… Can you elaborate on your algorithyms? Key is to have both algorythims and social interaction with the user.

Search and discovery are not exclusive. How does search interplay with discovery?

Francisco.. “We really want to help user find things..”

Greenstein..”Search is the beginning of discovery”…

Martino… “More data allows you to disambiguate”

Francisco…”Uploaded video, is somewhat of a face to face conversation…”

Hyman…”We put on everything we can to put safety in place…”   Music tastes bind people together.

Not sure that I have stopped to think about discovery, but maybe I will…